Datasets:

ICLR2024

/

ICLR2024-papers

like 2

Follow

ICLR2024 5

Poster

We present the Evolving Graph Fourier Transform (EFT), the first invertible spectral transform that captures evolving representations on temporal graphs. We motivate our work by the inadequacy of existing methods for capturing the evolving graph spectra, which are also computationally expensive due to the temporal aspect along with the graph vertex domain. We view the problem as an optimization over the Laplacian of the continuous time dynamic graph. Additionally, we propose pseudo-spectrum relaxations that decompose the transformation process, making it highly computationally efficient. The EFT method adeptly captures the evolving graph's structural and positional properties, making it effective for downstream tasks on evolving graphs. Hence, as a reference implementation, we develop a simple neural model induced with \eft for capturing evolving graph spectra. We empirically validate our theoretical findings on a number of large-scale and standard temporal graph benchmarks and demonstrate that our model achieves state-of-the-art performance.

Beyond Spatio-Temporal Representations: Evolving Fourier Transform for Temporal Graphs

2402.16078

https://openreview.net/forum?id=uvFhCUPjtI

Poster

Because diffusion models have shown impressive performances in a number of tasks, such as image synthesis, there is a trend in recent works to prove (with certain assumptions) that these models have strong approximation capabilities. In this paper, we show that current diffusion models actually have an expressive bottleneck in backward denoising and some assumption made by existing theoretical guarantees is too strong. Based on this finding, we prove that diffusion models have unbounded errors in both local and global denoising. In light of our theoretical studies, we introduce soft mixture denoising (SMD), an expressive and efficient model for backward denoising. SMD not only permits diffusion models to well approximate any Gaussian mixture distributions in theory, but also is simple and efficient for implementation. Our experiments on multiple image datasets show that SMD significantly improves different types of diffusion models (e.g., DDPM), espeically in the situation of few backward iterations.

Soft Mixture Denoising: Beyond the Expressive Bottleneck of Diffusion Models

2309.14068

https://openreview.net/forum?id=aaBnFAyW9O

Poster

Inducing and leveraging sparse activations during training and inference is a promising avenue for improving the computational efficiency of deep networks, which is increasingly important as network sizes continue to grow and their application becomes more widespread. Here we use the large width Gaussian process limit to analyze the behaviour, at random initialization, of nonlinear activations that induce sparsity in the hidden outputs. A previously unreported form of training instability is proven for arguably two of the most natural candidates for hidden layer sparsification; those being a shifted ReLU ($\phi(x)=\max(0, x-\tau)$ for $\tau\ge 0$) and soft thresholding ($\phi(x)=0$ for $|x|\le\tau$ and $x-\text{sign}(x)\tau$ for $|x|>\tau$). We show that this instability is overcome by clipping the nonlinear activation magnitude, at a level prescribed by the shape of the associated Gaussian process variance map. Numerical experiments verify the theory and show that the proposed magnitude clipped sparsifying activations can be trained with training and test fractional sparsity as high as 85\% while retaining close to full accuracy.

DEEP NEURAL NETWORK INITIALIZATION WITH SPARSITY INDUCING ACTIVATIONS

2402.16184

https://openreview.net/forum?id=uvXK8Xk9Jk

Poster

Treatment effect estimation in continuous time is crucial for personalized medicine. However, existing methods for this task are limited to point estimates of the potential outcomes, whereas uncertainty estimates have been ignored. Needless to say, uncertainty quantification is crucial for reliable decision-making in medical applications. To fill this gap, we propose a novel Bayesian neural controlled differential equation (BNCDE) for treatment effect estimation in continuous time. In our BNCDE, the time dimension is modeled through a coupled system of neural controlled differential equations and neural stochastic differential equations, where the neural stochastic differential equations allow for tractable variational Bayesian inference. Thereby, for an assigned sequence of treatments, our BNCDE provides meaningful posterior predictive distributions of the potential outcomes. To the best of our knowledge, ours is the first tailored neural method to provide uncertainty estimates of treatment effects in continuous time. As such, our method is of direct practical value for promoting reliable decision-making in medicine.

Bayesian Neural Controlled Differential Equations for Treatment Effect Estimation

2310.17463

https://openreview.net/forum?id=uwO71a8wET

Poster

Proximal policy optimization (PPO) has gained popularity in reinforcement learning (RL). Its PPO-Clip variant is one the most frequently implemented algorithms and is one of the first-to-try algorithms in RL tasks. This variant uses a clipped surrogate objective function not typically found in other algorithms. Many works have demonstrated the practical performance of PPO-Clip, but the theoretical understanding of it is limited to specific settings. In this work, we provide a comprehensive analysis that shows the stationary point convergence of PPO-Clip and the convergence rate thereof. Our analysis is new and overcomes many challenges, including the non-smooth nature of the clip operator, the potentially unbounded score function, and the involvement of the ratio of two stochastic policies. Our results and techniques might share new insights into PPO-Clip.

On Stationary Point Convergence of PPO-Clip

https://openreview.net/forum?id=uznKlCpWjV

Spotlight Poster

In the classical transformer attention scheme, we are given three $n \times d$ size matrices $Q, K, V$ (the query, key, and value tokens), and the goal is to compute a new $n \times d$ size matrix $D^{-1} \exp(QK^\top) V$ where $D = \mathrm{diag}( \exp(QK^\top) {\bf 1}_n )$. Here, $\exp()$ is applied entry-wise and ${\bf 1}_n$ denotes a length-$n$ vector whose entries are all ones.Intuitively, attention computation captures pairwise information between words in a sentence, but not higher-order information. Indeed, recent work \cite{sht23} has shown that attention units cannot solve simple problems about detecting triples of connected words.In this work, we study a generalization of attention which captures triple-wise correlations. The generalization is based on computations involving tensors defined by tuples of words. More formally, given five $n \times d$ size matrices $Q, K_1, K_2, V_1$ and $V_2$ (generalized query, key, and value tokens), our new goal is to compute an $n \times d$ size matrix $D^{-1} \exp( Q ( K_1 \oslash K_2)^\top ) (V_1 \oslash V_2) $ where $D = \mathrm{diag}( \exp( Q ( K_1 \oslash K_2)^\top ) {\bf 1}_{n^2} )$ and $K_1 \oslash K_2 \in \mathbb{R}^{n^2 \times d}$ denotes the column-wise Kronecker product of $K_1$ and $K_2$. This generalization is indeed able to solve problems about detecting triple-wise connections that were shown to be impossible for transformers.The potential downside of this generalization is that it appears as though computations are even more difficult, since the straightforward algorithm requires cubic time in $n$. However, we show that in the bounded-entry setting (which arises in practice, and which is well-studied in both theory and practice), there is actually a near-linear time algorithm. More precisely, we show that bounded entries are both necessary and sufficient for quickly performing generalized computations:$\bullet$ On the positive side, if all entries of the input matrices are bounded above by $o(\sqrt[3]{\log n})$ then we show how to approximate the ``tensor-type'' attention matrix in $n^{1+o(1)}$ time.$\bullet$ On the negative side, we show that if the entries of the input matrices may be as large as $\Omega(\sqrt[3]{\log n})$, then there is no algorithm that runs faster than $n^{3-o(1)}$ (assuming the Strong Exponential Time Hypothesis from fine-grained complexity theory).We also show that our construction, algorithms, and lower bounds naturally generalize to higher-order tensors and correlations. Interestingly, the higher the order of the tensors, the lower the bound on the entries needs to be for an efficient algorithm. Our results thus yield a natural tradeoff between the boundedness of the entries, and order of the tensor one may use for more expressive, efficient attention computation.Our constructions make use of a novel connection with a higher-order variant on the kernel density estimation problem. They combine a number of technical tools, including the polynomial method, algebraic geometry codes, and multiparty Merlin-Arthur communication protocols.

How to Capture Higher-order Correlations? Generalizing Matrix Softmax Attention to Kronecker Computation

2310.04064

https://openreview.net/forum?id=v0zNCwwkaV

Poster

Answering counterfactual queries has many important applications such as knowledge discovery and explainability, but is challenging when causal variables are unobserved and we only see a projection onto an observation space, for instance, image pixels. One approach is to recover the latent Structural Causal Model (SCM), but this typically needs unrealistic assumptions, such as linearity of the causal mechanisms. Another approach is to use naïve ML approximations, such as generative models, to generate counterfactual samples; however, these lack guarantees of accuracy. In this work, we strive to strike a balance between practicality and theoretical guarantees by focusing on a specific type of causal query called *domain counterfactuals*, which hypothesizes what a sample would have looked like if it had been generated in a different domain (or environment). Concretely, by only assuming invertibility, sparse domain interventions and access to observational data from different domains, we aim to improve domain counterfactual estimation both theoretically and practically with less restrictive assumptions. We define *domain counterfactually equivalent* models and prove necessary and sufficient properties for equivalent models that provide a tight characterization of the domain counterfactual equivalence classes. Building upon this result, we prove that every equivalence class contains a model where all intervened variables are at the end when topologically sorted by the causal DAG, i.e., all non-intervened variables have non-intervened ancestors. This surprising result suggests that a model design that only allows intervention in the last $k$ latent variables may improve model estimation for counterfactuals. We then test this model design on extensive simulated and image-based experiments which show the sparse canonical model indeed improves counterfactual estimation over baseline non-sparse models.

Towards Characterizing Domain Counterfactuals for Invertible Latent Causal Models

2306.11281

https://openreview.net/forum?id=v1VvCWJAL8

Poster

We propose Reinforcement Learning from Contrastive Distillation (RLCD), a method for aligning language models to follow principles expressed in natural language (e.g., to be more harmless) without using human feedback. RLCD creates preference pairs from two contrasting model outputs, one using a positive prompt designed to encourage following the given principles, and one using a negative prompt designed to encourage violating them. Using two different prompts causes model outputs to be more differentiated on average, resulting in cleaner preference labels in the absence of human annotations. We then use the preference pairs to train a preference model, which is in turn used to improve a base unaligned language model via reinforcement learning. Empirically, RLCD outperforms RLAIF (Bai et al., 2022b) and context distillation (Huang et al., 2022) baselines across three diverse alignment tasks—harmlessness, helpfulness, and story outline generation—and when using both 7B and 30B model scales for simulating preference data

RLCD: Reinforcement Learning from Contrastive Distillation for LM Alignment

https://openreview.net/forum?id=v3XXtxWKi6

Poster

We consider the problem of sampling from a logconcave distribution $\pi(\theta) \propto e^{-f(\theta)}$ constrained to a polytope $K:=${$\theta \in \mathbb{R}^d: A\theta \leq b$}, where $A\in \mathbb{R}^{m\times d}$ and $b \in \mathbb{R}^m$. The fastest-known algorithm for the setting when $f$ is $O(1)$-Lipschitz or $O(1)$-smooth runs in roughly $O(md \times md^{\omega -1})$ arithmetic operations, where the $md^{\omega -1}$ term arises because each Markov chain step requires computing a matrix inversion and determinant ($\omega \approx 2.37$ is the matrix multiplication constant). We present a nearly-optimal implementation of this Markov chain with per-step complexity that is roughly the number of non-zero entries of $A$ while the number of Markov chain steps remains the same. The key technical ingredients are 1) to show that the matrices that arise in this Dikin walk change slowly, 2) to deploy efficient linear solvers which can leverage this slow change to speed up matrix inversion by using information computed in previous steps, and 3) to speed up the computation of the determinantal term in the Metropolis filter step via a randomized Taylor series-based estimator. This result directly improves the runtime for applications that involve sampling from Gibbs distributions constrained to polytopes that arise in Bayesian statistics and private optimization.

Faster Sampling from Log-Concave Densities over Polytopes via Efficient Linear Solvers

https://openreview.net/forum?id=v63GWletn8

Poster

Score-based generative models like the diffusion model have been testified to be effective in modeling multi-modal data from image generation to reinforcement learning (RL). However, the inference process of diffusion model can be slow, which hinders its usage in RL with iterative sampling. We propose to apply the consistency model as an efficient yet expressive policy representation, namely consistency policy, with an actor-critic style algorithm for three typical RL settings: offline, offline-to-online and online. For offline RL, we demonstrate the expressiveness of generative models as policies from multi-modal data. For offline-to-online RL, the consistency policy is shown to be more computational efficient than diffusion policy, with a comparable performance. For online RL, the consistency policy demonstrates significant speedup and even higher average performances than the diffusion policy.

Consistency Models as a Rich and Efficient Policy Class for Reinforcement Learning

2309.16984

https://openreview.net/forum?id=v8jdwkUNXb

Poster

How humans and machines make sense of current inputs for relation reasoning and question-answering, and put the perceived information into context of our past memories, has been a challenging conundrum in cognitive science and artificial intelligence. Inspired by human brain's memory system and cognitive architectures, we propose a PMI framework that consists of perception, memory and inference components. Notably, the memory module comprises working and long-term memory, with the latter endowed with a higher-order structure to retain more accumulated knowledge and experiences. Through a differentiable competitive write access, current perceptions update working memory, which is later merged with long-term memory via outer product associations, averting memory overflow and minimizing information conflicts. In the inference module, relevant information is retrieved from two separate memory origins and associatively integrated to attain a more comprehensive and precise interpretation of current perceptions. We exploratively apply our PMI to improve prevailing Transformers and CNN models on question-answering tasks like bAbI-20k and Sort-of-CLEVR datasets, as well as relation reasoning and image classification tasks, and in each case, our PMI enhancements consistently outshine their original counterparts significantly. Visualization analyses reveal that memory consolidation, along with the interaction and integration of information from diverse memory sources, substantially contributes to the model effectiveness on inference tasks.

A Framework for Inference Inspired by Human Memory Mechanisms

2310.09297

https://openreview.net/forum?id=vBo7544jZx

Poster

Metagenomics studies genomic material derived from mixed microbial communities in diverse environments, holding considerable significance for both human health and environmental sustainability. Metagenomic binning refers to the clustering of genomic subsequences obtained from high-throughput DNA sequencing into distinct bins, each representing a constituent organism within the community. Mainstream binning methods primarily rely on sequence features such as composition and abundance, making them unable to effectively handle sequences shorter than 1,000 bp and inherent noise within sequences. Several binning tools have emerged, aiming to enhance binning outcomes by using the assembly graph generated by assemblers, which encodes valuable overlapping information among genomic sequences. However, existing assembly graph-based binners mainly focus on simplified contig-level assembly graphs that are recreated from assembler’s original graphs, unitig-level assembly graphs. The simplification reduces the resolution of the connectivity information in original graphs. In this paper, we design a novel binning tool named UnitigBin, which leverages representation learning on unitig-level assembly graphs while adhering to heterophilious constraints imposed by single-copy marker genes, ensuring that constrained contigs cannot be grouped together. Extensive experiments conducted on synthetic and real datasets demonstrate that UnitigBin significantly surpasses state-of-the-art binning tools.

Encoding Unitig-level Assembly Graphs with Heterophilous Constraints for Metagenomic Contigs Binning

https://openreview.net/forum?id=vBw8JGBJWj

Poster

In recent years, large language models have greatly improved in their ability to perform complex multi-step reasoning. However, even state-of-the-art models still regularly produce logical mistakes. To train more reliable models, we can turn either to outcome supervision, which provides feedback for a final result, or process supervision, which provides feedback for each intermediate reasoning step. Given the importance of training reliable models, and given the high cost of human feedback, it is important to carefully compare the both methods. Recent work has already begun this comparison, but many questions still remain. We conduct our own investigation, finding that process supervision significantly outperforms outcome supervision for training models to solve problems from the challenging MATH dataset. Our process-supervised model solves 78% of problems from a representative subset of the MATH test set. Additionally, we show that active learning significantly improves the efficacy of process supervision. To support related research, we also release PRM800K, the complete dataset of 800,000 step-level human feedback labels used to train our best reward model.

Let's Verify Step by Step

2305.20050

https://openreview.net/forum?id=v8L0pN6EOi

Poster

Existing losses used in deep metric learning (DML) for image retrieval often lead to highly non-uniform intra-class and inter-class representation structures across test classes and data distributions. When combined with the common practice of using a fixed threshold to declare a match, this gives rise to significant performance variations in terms of false accept rate (FAR) and false reject rate (FRR) across test classes and data distributions. We define this issue in DML as threshold inconsistency. In real-world applications, such inconsistency often complicates the threshold selection process when deploying large-scale image retrieval systems. To measure this inconsistency, we propose a novel variance-based metric called Operating-Point-Inconsistency-Score (OPIS) that quantifies the variance in the operating characteristics across classes. Using the OPIS metric, we find that achieving high accuracy levels in a DML model does not automatically guarantee threshold consistency. In fact, our investigation reveals a Pareto frontier in the high-accuracy regime, where existing methods to improve accuracy often lead to degradation in threshold consistency. To address this trade-off, we introduce the Threshold-Consistent Margin (TCM) loss, a simple yet effective regularization technique that promotes uniformity in representation structures across classes by selectively penalizing hard sample pairs. Large-scale experiments demonstrate TCM's effectiveness in enhancing threshold consistency while preserving accuracy, simplifying the threshold selection process in practical DML settings.

Threshold-Consistent Margin Loss for Open-World Deep Metric Learning

2307.04047

https://openreview.net/forum?id=vE5MyzpP92

Poster

While large language models (LLMs) exhibit remarkable capabilities across a wide range of tasks, they pose potential safety concerns, such as the ``jailbreak'' problem. Although several preventive measures have been developed to mitigate the potential risks associated with LLMs, they have primarily focused on English data. In this study, we reveal the presence of multilingual jailbreak challenges within LLMs and consider two potential risky scenarios: unintentional and intentional. The unintentional scenario involves users querying LLMs using non-English prompts and inadvertently bypassing the safety mechanisms, while the intentional scenario entails malicious users combining jailbreak instructions with multilingual prompts to attack LLMs deliberately. The experimental results reveal that in the unintentional scenario, the rate of unsafe content increases as the availability of languages decreases. Specifically, low-resource languages exhibit three times the likelihood of encountering harmful content compared to high-resource languages, with both ChatGPT and GPT-4. In the intentional scenario, multilingual prompts can exacerbate the negative impact of jailbreak instructions, with astonishingly high rates of unsafe output: 80.92\% for ChatGPT and 40.71\% for GPT-4. Finally, we propose a novel \textsc{Self-Defense} framework that addresses the multilingual jailbreak challenges via automatically generating multilingual safety training data for fine-tuning. Experiment results demonstrate its effectiveness with notable reduction in unsafe rate.

Multilingual Jailbreak Challenges in Large Language Models

2310.06474

https://openreview.net/forum?id=vESNKdEMGp

Poster

Motivated by equilibrium models of labor markets, we develop a formulation of causal strategic classification in which strategic agents can directly manipulate their outcomes. As an application, we consider employers that seek to anticipate the strategic response of a labor force when developing a hiring policy. We show theoretically that employers with performatively optimal hiring policies improve employer reward, labor force skill level, and labor force equity (compared to employers that do not anticipate the strategic labor force response) in the classic Coate-Loury labor market model. Empirically, we show that these desirable properties of performative hiring policies do generalize to our own formulation of a general equilibrium labor market. On the other hand, we also observe that the benefits of performatively optimal hiring policies are brittle in some aspects. We demonstrate that in our formulation a performative employer both harms workers by reducing their aggregate welfare and fails to prevent discrimination when more sophisticated wage and cost structures are introduced.

Learning in reverse causal strategic environments with ramifications on two sided markets

2404.13240

https://openreview.net/forum?id=vEfmVS5ywF

Poster

This work aims to improve the efficiency of vision transformers (ViTs). While ViTs use computationally expensive self-attention operations in every layer, we identify that these operations are highly correlated across layers -- a key redundancy that causes unnecessary computations. Based on this observation, we propose SkipAT a method to reuse self-attention computation from preceding layers to approximate attention at one or more subsequent layers. To ensure that reusing self-attention blocks across layers does not degrade the performance, we introduce a simple parametric function, which outperforms the baseline transformer's performance while running computationally faster. We show that SkipAT is agnostic to transformer architecture and is effective in image classification, semantic segmentation on ADE20K, image denoising on SIDD, and video denoising on DAVIS. We achieve improved throughput at the same-or-higher accuracy levels in all these tasks.

Skip-Attention: Improving Vision Transformers by Paying Less Attention

https://openreview.net/forum?id=vI95kcLAoU

Poster

Data attribution seeks to trace model outputs back to training data. With the recent development of diffusion models, data attribution has become a desired module to properly assign valuations for high-quality or copyrighted training samples, ensuring that data contributors are fairly compensated or credited. Several theoretically motivated methods have been proposed to implement data attribution, in an effort to improve the trade-off between computational scalability and effectiveness. In this work, we conduct extensive experiments and ablation studies on attributing diffusion models, specifically focusing on DDPMs trained on CIFAR-10 and CelebA, as well as a Stable Diffusion model LoRA-finetuned on ArtBench. Intriguingly, we report counter-intuitive observations that theoretically unjustified design choices for attribution empirically outperform previous baselines by a large margin, in terms of both linear datamodeling score and counterfactual evaluation. Our work presents a significantly more efficient approach for attributing diffusion models, while the unexpected findings suggest that at least in non-convex settings, constructions guided by theoretical assumptions may lead to inferior attribution performance.

Intriguing Properties of Data Attribution on Diffusion Models

2311.00500

https://openreview.net/forum?id=vKViCoKGcB

Poster

Recent Meta-learning for Black-Box Optimization (MetaBBO) methods harness neural networks to meta-learn configurations of traditional black-box optimizers. Despite their success, they are inevitably restricted by the limitations of predefined hand-crafted optimizers. In this paper, we present SYMBOL, a novel framework that promotes the automated discovery of black-box optimizers through symbolic equation learning. Specifically, we propose a Symbolic Equation Generator (SEG) that allows closed-form optimization rules to be dynamically generated for specific tasks and optimization steps. Within SYMBOL, we then develop three distinct strategies based on reinforcement learning, so as to meta-learn the SEG efficiently. Extensive experiments reveal that the optimizers generated by SYMBOL not only surpass the state-of-the-art BBO and MetaBBO baselines, but also exhibit exceptional zero-shot generalization abilities across entirely unseen tasks with different problem dimensions, population sizes, and optimization horizons. Furthermore, we conduct in-depth analyses of our SYMBOL framework and the optimization rules that it generates, underscoring its desirable flexibility and interpretability.

SYMBOL: Generating Flexible Black-Box Optimizers through Symbolic Equation Learning

2402.02355

https://openreview.net/forum?id=vLJcd43U7a

Poster

The thriving field of multi-agent reinforcement learning (MARL) studies how a group of interacting agents make decisions autonomously in a shared dynamic environment. Existing theoretical studies in this area suffer from at least two of the following obstacles: memory inefficiency, the heavy dependence of sample complexity on the long horizon and the large state space, the high computational complexity, non-Markov policy, non-Nash policy, and high burn-in cost. In this work, we take a step towards settling this problem by designing a model-free self-play algorithm \emph{Memory-Efficient Nash Q-Learning (ME-Nash-QL)} for two-player zero-sum Markov games, which is a specific setting of MARL. We prove that ME-Nash-QL can output an $\varepsilon$-approximate Nash policy with remarkable space complexity $O(SABH)$, sample complexity $\widetilde{O}(H^4SAB/\varepsilon^2)$, and computational complexity $O(T\mathrm{poly}(AB))$, where $S$ is the number of states, $\{A, B\}$ is the number of actions for the two players, $H$ is the horizon length, and $T$ is the number of samples. Notably, our approach outperforms in terms of space complexity compared to existing algorithms for tabular cases. It achieves the lowest computational complexity while preserving Markov policies, setting a new standard. Furthermore, our algorithm outputs a Nash policy and achieves the best sample complexity compared with the existing guarantee for long horizons, i.e. when $\min \\{ A, B \\} \ll H^2$. Our algorithm also achieves the best burn-in cost $O(SAB\,\mathrm{poly}(H))$, whereas previous algorithms need at least $O(S^3 AB\,\mathrm{poly}(H))$ to attain the same level of sample complexity with ours.

Provable Memory Efficient Self-Play Algorithm for Model-free Reinforcement Learning

https://openreview.net/forum?id=vNiI3aGcE6

Poster

Recent deep neural networks (DNNs) have came to rely on vast amounts of training data, providing an opportunity for malicious attackers to exploit and contaminate the data to carry out backdoor attacks. However, existing backdoor attack methods make unrealistic assumptions, assuming that all training data comes from a single source and that attackers have full access to the training data. In this paper, we introduce a more realistic attack scenario where victims collect data from multiple sources, and attackers cannot access the complete training data. We refer to this scenario as $\textbf{data-constrained backdoor attacks}$. In such cases, previous attack methods suffer from severe efficiency degradation due to the $\textbf{entanglement}$ between benign and poisoning features during the backdoor injection process. To tackle this problem, we introduce three CLIP-based technologies from two distinct streams: $\textit{Clean Feature Suppression}$ and $\textit{Poisoning Feature Augmentation}$. The results demonstrate remarkable improvements, with some settings achieving over $\textbf{100}$% improvement compared to existing attacks in data-constrained scenarios.

Efficient Backdoor Attacks for Deep Neural Networks in Real-world Scenarios

2306.08386

https://openreview.net/forum?id=vRyp2dhEQp

Spotlight Poster

Transformers pretrained on diverse tasks exhibit remarkable in-context learning (ICL) capabilities, enabling them to solve unseen tasks solely based on input contexts without adjusting model parameters. In this paper, we study ICL in one of its simplest setups: pretraining a single-layer linear attention model for linear regression with a Gaussian prior. We establish a statistical task complexity bound for the attention model pretraining, showing that effective pretraining only requires a small number of independent tasks. Furthermore, we prove that the pretrained model closely matches the Bayes optimal algorithm, i.e., optimally tuned ridge regression, by achieving nearly Bayes optimal risk on unseen tasks under a fixed context length. These theoretical findings complement prior experimental research and shed light on the statistical foundations of ICL.

How Many Pretraining Tasks Are Needed for In-Context Learning of Linear Regression?

2310.08391

https://openreview.net/forum?id=vSh5ePa0ph

Poster

Transfer learning has recently shown significant performance across various tasks involving deep neural networks. In these transfer learning scenarios, the prior distribution for downstream data becomes crucial in Bayesian model averaging (BMA). While previous works proposed the prior over the neural network parameters centered around the pre-trained solution, such strategies have limitations when dealing with distribution shifts between upstream and downstream data. This paper introduces nonparametric transfer learning (NPTL), a flexible posterior sampling method to address the distribution shift issue within the context of nonparametric learning. The nonparametric learning (NPL) method is a recent approach that employs a nonparametric prior for posterior sampling, efficiently accounting for model misspecification scenarios, which is suitable for transfer learning scenarios that may involve the distribution shift between upstream and downstream tasks. Through extensive empirical validations, we demonstrate that our approach surpasses other baselines in BMA performance.

Enhancing Transfer Learning with Flexible Nonparametric Posterior Sampling

2403.07282

https://openreview.net/forum?id=vSwu81S33z

Poster

In this paper, we present a hybrid X-shaped vision Transformer, named Xformer, which performs notably on image denoising tasks. We explore strengthening the global representation of tokens from different scopes. In detail, we adopt two types of Transformer blocks. The spatial-wise Transformer block performs fine-grained local patches interactions across tokens defined by spatial dimension. The channel-wise Transformer block performs direct global context interactions across tokens defined by channel dimension. Based on the concurrent network structure, we design two branches to conduct these two interaction fashions. Within each branch, we employ an encoder-decoder architecture to capture multi-scale features. Besides, we propose the Bidirectional Connection Unit (BCU) to couple the learned representations from these two branches while providing enhanced information fusion. The joint designs make our Xformer powerful to conduct global information modeling in both spatial and channel dimensions. Extensive experiments show that Xformer, under the comparable model complexity, achieves state-of-the-art performance on the synthetic and real-world image denoising tasks.

Xformer: Hybrid X-Shaped Transformer for Image Denoising

https://openreview.net/forum?id=vXrIQLzIKY

Poster

Large language models have become the cornerstone of natural language processing, but their use comes with substantial costs in terms of compute and memory resources. Sparsification provides a solution to alleviate these resource constraints, and recent works have shown that trained models can be sparsified post-hoc. Existing sparsification techniques face challenges as they need additional data structures and offer constrained speedup with current hardware. In this paper we present SliceGPT, a new post-training sparsification scheme which replaces each weight matrix with a smaller (dense) matrix, reducing the embedding dimension of the network. Through extensive experimentation, we show that SliceGPT can remove up to 25\% of the model parameters (including embeddings) for OPT 66B and LLAMA-2 70B models with negligible loss in accuracy. Our sliced models run on fewer GPUs and run faster without any additional code optimization: on 24GB consumer GPUs we reduce the total compute for inference on LLAMA-2 70B to 64% of that of the dense model; on 40GB A100 GPUs we reduce it to 66%. We offer a new insight, computational invariance in transformer networks, which enables SliceGPT and we hope it will inspire and enable future avenues to reduce memory and computation demands for pre-trained models.

SliceGPT: Compress Large Language Models by Deleting Rows and Columns

2401.15024

https://openreview.net/forum?id=vXxardq6db

Poster

Recent advancements in neural vocoding are predominantly driven by Generative Adversarial Networks (GANs) operating in the time-domain. While effective, this approach neglects the inductive bias offered by time-frequency representations, resulting in reduntant and computionally-intensive upsampling operations. Fourier-based time-frequency representation is an appealing alternative, aligning more accurately with human auditory perception, and benefitting from well-established fast algorithms for its computation. Nevertheless, direct reconstruction of complex-valued spectrograms has been historically problematic, primarily due to phase recovery issues. This study seeks to close this gap by presenting Vocos, a new model that directly generates Fourier spectral coefficients. Vocos not only matches the state-of-the-art in audio quality, as demonstrated in our evaluations, but it also substantially improves computational efficiency, achieving an order of magnitude increase in speed compared to prevailing time-domain neural vocoding approaches. The source code and model weights have been open-sourced.

Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis

2306.00814

https://openreview.net/forum?id=vY9nzQmQBw

Poster

Large Language Models (LLMs) have already become quite proficient at solving simpler programming tasks like those in HumanEval or MBPP benchmarks. However, solving more complex and competitive programming tasks is still quite challenging for these models - possibly due to their tendency to generate solutions as monolithic code blocks instead of decomposing them into logical sub-tasks and sub-modules. On the other hand, experienced programmers instinctively write modularized code with abstraction for solving complex tasks, often reusing previously developed modules. To address this gap, we propose CodeChain, a novel framework for inference that elicits modularized code generation through a chain of self-revisions, each being guided by some representative sub-modules generated in previous iterations. Concretely, CodeChain first instructs the LLM to generate modularized codes through chain-of-thought prompting. Then it applies a chain of self-revisions by iterating the two steps: 1) extracting and clustering the generated sub-modules and selecting the cluster representatives as the more generic and re-usable implementations, and 2) augmenting the original chain-of-thought prompt with these selected module-implementations and instructing the LLM to re-generate new modularized solutions. We find that by naturally encouraging the LLM to reuse the previously developed and verified sub-modules, CodeChain can significantly boost both modularity as well as correctness of the generated solutions, achieving relative pass@1 improvements of 35\% on APPS and 76\% on CodeContests. It is shown to be effective on both OpenAI LLMs as well as open-sourced LLMs like WizardCoder. We also conduct comprehensive ablation studies with different methods of prompting, number of clusters, model sizes, program qualities, etc., to provide useful insights that underpin CodeChain's success.

CodeChain: Towards Modular Code Generation Through Chain of Self-revisions with Representative Sub-modules

2310.08992

https://openreview.net/forum?id=vYhglxSj8j

Poster

Recent works have shown that deep neural networks are vulnerable to adversarial examples that find samples close to the original image but can make the model misclassify. Even with access only to the model's output, an attacker can employ black-box attacks to generate such adversarial examples. In this work, we propose a simple and lightweight defense against black-box attacks by adding random noise to hidden features at intermediate layers of the model at inference time. Our theoretical analysis confirms that this method effectively enhances the model's resilience against both score-based and decision-based black-box attacks. Importantly, our defense does not necessitate adversarial training and has minimal impact on accuracy, rendering it applicable to any pre-trained model. Our analysis also reveals the significance of selectively adding noise to different parts of the model based on the gradient of the adversarial objective function, which can be varied during the attack. We demonstrate the robustness of our defense against multiple black-box attacks through extensive empirical experiments involving diverse models with various architectures.

Understanding the Robustness of Randomized Feature Defense Against Query-Based Adversarial Attacks

2310.00567

https://openreview.net/forum?id=vZ6r9GMT1n

Poster

Communication is a powerful tool for coordination in multi-agent RL. But inducing an effective, common language is a difficult challenge, particularly in the decentralized setting. In this work, we introduce an alternative perspective where communicative messages sent between agents are considered as different incomplete views of the environment state. By examining the relationship between messages sent and received, we propose to learn to communicate using contrastive learning to maximize the mutual information between messages of a given trajectory. In communication-essential environments, our method outperforms previous work in both performance and learning speed. Using qualitative metrics and representation probing, we show that our method induces more symmetric communication and captures global state information from the environment. Overall, we show the power of contrastive learning and the importance of leveraging messages as encodings for effective communication.

Learning Multi-Agent Communication with Contrastive Learning

2307.01403

https://openreview.net/forum?id=vZZ4hhniJU

Poster

Wide networks usually yield better accuracy than their narrower counterpart at the expense of the massive $\texttt{mult}$ cost.To break this tradeoff, we advocate a novel concept of $\textit{Structured Activation Sparsification}$, dubbed SAS, which boosts accuracy without increasing computation by utilizing the projected sparsity in activation maps with a specific structure. Concretely, the projected sparse activation is allowed to have N nonzero value among M consecutive activations.Owing to the local structure in sparsity, the wide $\texttt{matmul}$ between a dense weight and the sparse activation is executed as an equivalent narrow $\texttt{matmul}$ between a dense weight and dense activation, which is compatible with NVIDIA's $\textit{SparseTensorCore}$ developed for the N:M structured sparse weight.In extensive experiments, we demonstrate that increasing sparsity monotonically improves accuracy (up to 7% on CIFAR10) without increasing the $\texttt{mult}$ count.Furthermore, we show that structured sparsification of $\textit{activation}$ scales better than that of $\textit{weight}$ given the same computational budget.

SAS: Structured Activation Sparsification

https://openreview.net/forum?id=vZfi5to2Xl

Poster

The conventional deep learning paradigm often involves training a deep model on a server and then deploying the model or its distilled ones to resource-limited edge devices. Usually, the models shall remain fixed once deployed (at least for some period) due to the potential high cost of model adaptation for both the server and edge sides. However, in many real-world scenarios, the test environments may change dynamically (known as distribution shifts), which often results in degraded performance. Thus, one has to adapt the edge models promptly to attain promising performance. Moreover, with the increasing data collected at the edge, this paradigm also fails to further adapt the cloud model for better performance. To address these, we encounter two primary challenges: 1) the edge model has limited computation power and may only support forward propagation; 2) the data transmission budget between cloud and edge devices is limited in latency-sensitive scenarios. In this paper, we establish a Cloud-Edge Model Adaptation (CEMA) paradigm in which the edge models only need to perform forward propagation and the edge models can be adapted online. In our CEMA, to reduce the communication burden, we devise two criteria to exclude unnecessary samples from uploading to the cloud, i.e., dynamic unreliable and low-informative sample exclusion. Based on the uploaded samples, we update and distribute the affine parameters of normalization layers by distilling from the stronger foundation model to the edge model with a sample replay strategy. Extensive experimental results on ImageNet-C and ImageNet-R verify the effectiveness of our CEMA.

Towards Robust and Efficient Cloud-Edge Elastic Model Adaptation via Selective Entropy Distillation

2402.17316

https://openreview.net/forum?id=vePdNU3u6n

Poster

We propose a novel random walk-based algorithm for unbiased estimation of arbitrary functions of a weighted adjacency matrix, coined universal graph random features (u-GRFs). This includes many of the most popular examples of kernels defined on the nodes of a graph. Our algorithm enjoys subquadratic time complexity with respect to the number of nodes, overcoming the notoriously prohibitive cubic scaling of exact graph kernel evaluation. It can also be trivially distributed across machines, permitting learning on much larger networks. At the heart of the algorithm is a modulation function which upweights or downweights the contribution from different random walks depending on their lengths. We show that by parameterising it with a neural network we can obtain u-GRFs that give higher-quality kernel estimates or perform efficient, scalable kernel learning. We provide robust theoretical analysis and support our findings with experiments including pointwise estimation of fixed graph kernels, solving non-homogeneous graph ordinary differential equations, node clustering and kernel regression on triangular meshes.

General Graph Random Features

2310.04859

https://openreview.net/forum?id=viftsX50Rt

Poster

For object re-identification (re-ID), learning from synthetic data has become a promising strategy to cheaply acquire large-scale annotated datasets and effective models, with few privacy concerns. Many interesting research problems arise from this strategy, e.g., how to reduce the domain gap between synthetic source and real-world target. To facilitate developing more new approaches in learning from synthetic data, we introduce the Alice benchmarks, large-scale datasets providing benchmarks as well as evaluation protocols to the research community. Within the Alice benchmarks, two object re-ID tasks are offered: person and vehicle re-ID. We collected and annotated two challenging real-world target datasets: AlicePerson and AliceVehicle, captured under various illuminations, image resolutions, etc. As an important feature of our real target, the clusterability of its training set is not manually guaranteed to make it closer to a real domain adaptation test scenario. Correspondingly, we reuse existing PersonX and VehicleX as synthetic source domains. The primary goal is to train models from synthetic data that can work effectively in the real world. In this paper, we detail the settings of Alice benchmarks, provide an analysis of existing commonly-used domain adaptation methods, and discuss some interesting future directions. An online server will be set up for the community to evaluate methods conveniently and fairly.

Alice Benchmarks: Connecting Real World Re-Identification with the Synthetic

2310.04416

https://openreview.net/forum?id=vkkHqoerLV

Spotlight Poster

Recently, Transformer-based and MLP-based models have emerged rapidly and won dominance in time series analysis. In contrast, convolution is losing steam in time series tasks nowadays for inferior performance. This paper studies the open question of how to better use convolution in time series analysis and makes efforts to bring convolution back to the arena of time series analysis. To this end, we modernize the traditional TCN and conduct time series related modifications to make it more suitable for time series tasks. As the outcome, we propose ModernTCN and successfully solve this open question through a seldom-explored way in time series community. As a pure convolution structure, ModernTCN still achieves the consistent state-of-the-art performance on five mainstream time series analysis tasks (long-term and short-term forecasting, imputation, classification and anomaly detection) while maintaining the efficiency advantage of convolution-based models, therefore providing a better balance of efficiency and performance than state-of-the-art Transformer-based and MLP-based models. Our study further reveals that, compared with previous convolution-based models, our ModernTCN has much larger effective receptive fields (ERFs), therefore can better unleash the potential of convolution in time series analysis. The code will be publicly available.

ModernTCN: A Modern Pure Convolution Structure for General Time Series Analysis

https://openreview.net/forum?id=vpJMJerXHU

Spotlight Poster

It has been observed that representations learned by distinct neural networks conceal structural similarities when the models are trained under similar inductive biases. From a geometric perspective, identifying the classes of transformations and the related invariances that connect these representations is fundamental to unlocking applications, such as merging, stitching, and reusing different neural modules. However, estimating task-specific transformations a priori can be challenging and expensive due to several factors (e.g., weights initialization, training hyperparameters, or data modality). To this end, we introduce a versatile method to directly incorporate a set of invariances into the representations, constructing a product space of invariant components on top of the latent representations without requiring prior knowledge about the optimal invariance to infuse. We validate our solution on classification and reconstruction tasks, observing consistent latent similarity and downstream performance improvements in a zero-shot stitching setting. The experimental analysis comprises three modalities (vision, text, and graphs), twelve pretrained foundational models, eight benchmarks, and several architectures trained from scratch.

From Bricks to Bridges: Product of Invariances to Enhance Latent Space Communication

2310.01211

https://openreview.net/forum?id=vngVydDWft

Poster

Offline reinforcement learning (RL) allows agents to learn effective, return-maximizing policies from a static dataset. Three popular algorithms for offline RL are Conservative Q-Learning (CQL), Behaviour Cloning (BC), and Decision Transformer (DT), from the class of Q-Learning, Imitation Learning, and Sequence Modeling respectively. A key open question is: which algorithm is preferred under what conditions? We study this question empirically by exploring the performance of these algorithms across the commonly used D4RL and Robomimic benchmarks. We design targeted experiments to understand their behavior concerning data suboptimality, task complexity, and stochasticity. Our key findings are: (1) DT requires more data than CQL to learn competitive policies but is more robust; (2) DT is a substantially better choice than both CQL and BC in sparse-reward and low-quality data settings; (3) DT and BC are preferable as task horizon increases, or when data is obtained from human demonstrators; and (4) CQL excels in situations characterized by the combination of high stochasticity and lower data quality. We also investigate architectural choices and scaling trends for DT on Atari and D4RL and make design/scaling recommendations. We find that scaling the amount of data for DT by 5x gives a 2.5x average score improvement on Atari.

When should we prefer Decision Transformers for Offline Reinforcement Learning?

2305.14550

https://openreview.net/forum?id=vpV7fOFQy4

Poster

Causal inference is one of the hallmarks of human intelligence. While the field of CausalNLP has attracted much interest in the recent years, existing causal inference datasets in NLP primarily rely on discovering causality from empirical knowledge (e.g. commonsense knowledge). In this work, we propose the first benchmark dataset to test the pure causal inference skills of large language models (LLMs). Specifically, we formulate a novel task Corr2Cause, which takes a set of correlational statements and determines the causal relationship between the variables. We curate a large-scale dataset of more than 400K samples, on which we evaluate seventeen existing LLMs. Through our experiments, we identify a key shortcoming of LLMs in terms of their causal inference skills, and show that these models achieve almost close to random performance on the task. This shortcoming is somewhat mitigated when we try to re-purpose LLMs for this skill via finetuning, but we find that these models still fail to generalize – they can only perform causal inference in in-distribution settings when variable names and textual expressions used in the queries are similar to those in the training set, but fail in out-of-distribution settings generated by perturbing these queries. Corr2Cause is a challenging task for LLMs, and would be helpful in guiding future research on improving LLMs’ pure reasoning ability and generalizability.

Can Large Language Models Infer Causation from Correlation?

2306.05836

https://openreview.net/forum?id=vqIH0ObdqL

Spotlight Poster

Estimating the properties of quantum systems such as quantum phase has been critical in addressing the essential quantum many-body problems in physics and chemistry. Deep learning models have been recently introduced to property estimation, surpassing conventional statistical approaches. However, these methods are tailored to the specific task and quantum data at hand. It remains an open and attractive question for devising a more universal task-agnostic pre-training model for quantum property estimation. In this paper, we propose LLM4QPE, a large language model style quantum task-agnostic pre-training and finetuning paradigm that 1) performs unsupervised pretraining on diverse quantum systems with different physical conditions; 2) uses the pretrained model for supervised finetuning and delivers high performance with limited training data, on downstream tasks. It mitigates the cost for quantum data collection and speeds up convergence. Extensive experiments show the promising efficacy of LLM4QPE in various tasks including classifying quantum phases of matter on Rydberg atom model and predicting two-body correlation function on anisotropic Heisenberg model.

Towards LLM4QPE: Unsupervised Pretraining of Quantum Property Estimation and A Benchmark

https://openreview.net/forum?id=vrBVFXwAmi

Poster

We study best arm identification (BAI) in linear bandits in the fixed-budget regime under differential privacy constraints, when the arm rewards are supported on the unit interval. Given a finite budget $T$ and a privacy parameter $\varepsilon>0$, the goal is to minimise the error probability in finding the arm with the largest mean after $T$ sampling rounds, subject to the constraint that the policy of the decision maker satisfies a certain {\em $\varepsilon$-differential privacy} ($\varepsilon$-DP) constraint. We construct a policy satisfying the $\varepsilon$-DP constraint (called {\sc DP-BAI}), based on the principle of {\em maximum absolute determinants}, and derive an upper bound on its error probability. Furthermore, we derive a minimax lower bound on the error probability, and demonstrate that the lower and the upper bounds decay exponentially in $T$, with exponents in the two bounds matching order-wise in (a) the sub-optimality gaps of the arms, (b) $\varepsilon$, and (c) the problem complexity that is expressible as the sum of two terms, one characterising the complexity of standard fixed-budget BAI (without privacy constraints), and the other accounting for the $\varepsilon$-DP constraint. Additionally, we present some auxiliary results that contribute to the derivation of the lower bound on the error probability. These results, we posit, may be of independent interest and could prove instrumental in proving lower bounds on error probabilities in several other bandit problems.Whereas prior works provide results for BAI in the fixed-budget regime without privacy constraints or in the fixed-confidence regime with privacy constraints, our work fills the gap in the literature by providing the results for BAI in the fixed-budget regime under the $\varepsilon$-DP constraint.

Fixed-Budget Differentially Private Best Arm Identification

2401.09073

https://openreview.net/forum?id=vrE2fqAInO

Poster

We propose that the grokking phenomenon, where the train loss of a neural network decreases much earlier than its test loss, can arise due to a neural network transitioning from lazy training dynamics to a rich, feature learning regime. To illustrate this mechanism, we study the simple setting of vanilla gradient descent on a polynomial regression problem with a two layer neural network which exhibits grokking without regularization in a way that cannot be explained by existing theories. We identify sufficient statistics for the test loss of such a network, and tracking these over training reveals that grokking arises in this setting when the network first attempts to fit a kernel regression solution with its initial features, followed by late-time feature learning where a generalizing solution is identified after train loss is already low. We find that the key determinants of grokking are the rate of feature learning---which can be controlled precisely by parameters that scale the network output---and the alignment of the initial features with the target function $y(x)$. We argue this delayed generalization arises when (1) the top eigenvectors of the initial neural tangent kernel and the task labels $y(x)$ are misaligned, but (2) the dataset size is large enough so that it is possible for the network to generalize eventually, but not so large that train loss perfectly tracks test loss at all epochs, and (3) the network begins training in the lazy regime so does not learn features immediately. We conclude with evidence that this transition from lazy (linear model) to rich training (feature learning) can control grokking in more general settings, like on MNIST, one-layer Transformers, and student-teacher networks.

Grokking as the transition from lazy to rich training dynamics

2310.06110

https://openreview.net/forum?id=vt5mnLVIVo

Poster

Graph Visualization, also known as Graph Drawing, aims to find geometric embeddings of graphs that optimize certain criteria. Stress is a widely used metric; stress is minimized when every pair of nodes is positioned at their shortest path distance. However, stress optimization presents computational challenges due to its inherent complexity and is usually solved using heuristics in practice. We introduce a scalable Graph Neural Network (GNN) based Graph Drawing framework with sub-quadratic runtime that can learn to optimize stress. Inspired by classical stress optimization techniques and force-directed layout algorithms, we create a coarsening hierarchy for the input graph. Beginning at the coarsest level, we iteratively refine and un-coarsen the layout, until we generate an embedding for the original graph. To enhance information propagation within the network, we propose a novel positional rewiring technique based on intermediate node positions. Our empirical evaluation demonstrates that the framework achieves state-of-the-art performance while remaining scalable.

CoRe-GD: A Hierarchical Framework for Scalable Graph Visualization with GNNs

https://openreview.net/forum?id=vtyasLn4RM

Poster

Path planning underpins various applications such as transportation, logistics, and robotics.Conventionally, path planning is formulated with explicit optimization objectives such as distance or time.However, real-world data reveals that user intentions are hard-to-model, suggesting a need for data-driven path planning that implicitly incorporates the complex user intentions.In this paper, we propose GDP, a diffusion-based model for end-to-end data-driven path planning.It effectively learns path patterns via a novel diffusion process that incorporates constraints from road networks, and plans paths as conditional path generation given the origin and destination as prior evidence.GDP is the first solution that bypasses the traditional search-based frameworks, a long-standing performance bottleneck in path planning.We validate the efficacy of GDP on two real-world datasets.Our GDP beats strong baselines by 14.2% ~ 43.5% and achieves state-of-the-art performances.

GRAPH-CONSTRAINED DIFFUSION FOR END-TO-END PATH PLANNING

https://openreview.net/forum?id=vuK8MhVtuu

Poster

Primal-dual safe RL methods commonly perform iterations between the primal update of the policy and the dual update of the Lagrange Multiplier. Such a training paradigm is highly susceptible to the error in cumulative cost estimation since this estimation serves as the key bond connecting the primal and dual update processes. We show that this problem causes significant underestimation of cost when using off-policy methods, leading to the failure to satisfy the safety constraint. To address this issue, we propose conservative policy optimization, which learns a policy in a constraint-satisfying area by considering the uncertainty in cost estimation. This improves constraint satisfaction but also potentially hinders reward maximization. We then introduce local policy convexification to help eliminate such suboptimality by gradually reducing the estimation uncertainty. We provide theoretical interpretations of the joint coupling effect of these two ingredients and further verify them by extensive experiments. Results on benchmark tasks show that our method not only achieves an asymptotic performance comparable to state-of-the-art on-policy methods while using much fewer samples, but also significantly reduces constraint violation during training. Our code is available at https://github.com/ZifanWu/CAL.

Off-Policy Primal-Dual Safe Reinforcement Learning

2401.14758

https://openreview.net/forum?id=vy42bYs1Wo

Poster

In the large language models era, it is imperative to measure and understand how gender biases present in the training data influence model behavior.Previous works construct benchmarks around known stereotypes (e.g., occupations) and demonstrate high levels of gender bias in large language models, raising serious concerns about models exhibiting undesirable behaviors.We expand on existing literature by asking the question: \textit{Do large language models still favor one gender over the other in non-stereotypical settings?}To tackle this question, we restrict language model evaluation to a \textit{neutral} subset, in which sentences are free of pronounced word-gender associations. After characterizing these associations in terms of pretraining data statistics,we use them to (1) create a new benchmark with low gender-word associations, and (2) repurpose popular benchmarks in the gendered pronoun setting | WinoBias and \Winogender |, removing pronounced gender-correlated words.Surprisingly, when testing $20+$ models (e.g., Llama-2, Pythia, and OPT) in the proposed benchmarks, we still detect critically high gender bias across all tested models. For instance, after adjusting for strong word-gender associations, we find that all models still exhibit clear gender preferences in about $60\%$-$95\%$ of the sentences, representing a small change (up to $5\%$) from the original \textit{stereotypical} setting.By demonstrating that measured bias is not necessarily due to the presence of highly gender-associated words, our work highlights important questions about bias evaluation as well as potentially underlying model biases.

Are Models Biased on Text without Gender-related Language?

2405.00588

https://openreview.net/forum?id=w1JanwReU6

Poster

Self-supervised learning (SSL) has recently emerged as a promising paradigm for training generalisable models on large-scale data in the fields of vision, text, and speech. Although SSL has been proven effective in speech and audio, its application to music audio has yet to be thoroughly explored. This is partially due to the distinctive challenges associated with modelling musical knowledge, particularly tonal and pitched characteristics of music.To address this research gap, we propose an acoustic **M**usic und**ER**standing model with large-scale self-supervised **T**raining (**MERT**), which incorporates teacher models to provide pseudo labels in the masked language modelling (MLM) style acoustic pre-training.In our exploration, we identified an effective combination of teacher models, which outperforms conventional speech and audio approaches in terms of performance. This combination includes an acoustic teacher based on Residual Vector Quantization - Variational AutoEncoder (RVQ-VAE) and a musical teacher based on the Constant-Q Transform (CQT). Furthermore, we explore a wide range of settings to overcome the instability in acoustic language model pre-training, which allows our designed paradigm to scale from 95M to 330M parameters.Experimental results indicate that our model can generalise and perform well on 14 music understanding tasks and attain state-of-the-art (SOTA) overall scores.

MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training

2306.00107

https://openreview.net/forum?id=w3YZ9MSlBu

Poster

Large language models (LLMs) have made significant advancements in various natural language processing tasks but face challenges such as hallucinations and integration of up-to-date knowledge, which is particularly critical for question answering (QA). While incorporating new information with the retrieval of relevant passages is a promising way to improve QA with LLMs, the existing methods often require additional fine-tuning which becomes infeasible with recent LLMs. Retrieval augmentation via prompting has the potential to address this limitation, but this direction has been limitedly explored. To this end, we design a simple yet effective framework to enhance open-domain QA (ODQA) with LLMs, based on the summarized retrieval (SuRe). SuRe helps LLMs predict more grounded answers, which are well-supported by the summarization of retrieved passages that could be viewed as an explicit rationale extracted from the retrieved passages. Specifically, SuRe first constructs summaries of the retrieved passages for each of the multiple answer candidates. Then, SuRe confirms the most plausible answer from the candidate set by evaluating the validity and ranking of the generated summaries. Experimental results on diverse ODQA benchmarks demonstrate the superiority of SuRe, with improvements of up to 4.4\% in exact match (EM) and 3.9\% in F1 score over standard prompting approaches. SuRe also can be integrated with a broad range of retrieval methods and LLMs. Finally, the generated summaries from SuRe show additional advantages to measure the importance of retrieved passages and serve as more preferred rationales by models and humans.

SuRe: Summarizing Retrievals using Answer Candidates for Open-domain QA of LLMs

2404.13081

https://openreview.net/forum?id=w4DW6qkRmt

Poster

Reinforcement learning (RL) presents a promising framework to learn policies through environment interaction, but often requires an infeasible amount of interaction data to solve complex tasks from sparse rewards. One direction includes augmenting RL with offline data demonstrating desired tasks, but past work often require a lot of high-quality demonstration data that is difficult to obtain, especially for domains such as robotics. Our approach consists of a reverse curriculum followed by a forward curriculum. Unique to our approach compared to past work is the ability to efficiently leverage more than one demonstration via a per-demonstration reverse curriculum generated via state resets. The result of our reverse curriculum is an initial policy that performs well on a narrow initial state distribution and helps overcome difficult exploration problems. A forward curriculum is then used to accelerate the training of the initial policy to perform well on the full initial state distribution of the task and improve demonstration and sample efficiency. We show how the combination of a reverse curriculum and forward curriculum in our method, RFCL, enables significant improvements in demonstration and sample efficiency compared against various state-of-the-art learning-from-demonstration baselines, even solving previously unsolvable tasks that require high precision and control. Website with code and visualizations are here: https://reverseforward-cl.github.io/

Reverse Forward Curriculum Learning for Extreme Sample and Demo Efficiency

https://openreview.net/forum?id=w4rODxXsmM

Poster

Interference is ubiquitous when conducting causal experiments over social networks. Except for certain network structures, causal inference on the network in the presence of interference is difficult due to the entanglement between the treatment assignments and the interference levels. In this article, we conduct causal inference under interference on an observed, sparse but connected network, and we propose a novel design of experiments based on an independent set. Compared to conventional designs, the independent-set design focuses on an independent subset of data and controls their interference exposures through the assignments to the rest (auxiliary set). The independent-set design enhances the performance of causal estimators by trading sample quantity for sample quality. We show the capacity of our approach for various causal inference tasks, justify its superiority over conventional methods, and illustrate the empirical performance through simulations.

Independent-Set Design of Experiments for Estimating Treatment and Spillover Effects under Network Interference

2312.04026

https://openreview.net/forum?id=w50MQ9Vfty

Poster

Estimating neural radiance fields (NeRFs) is able to generate novel views of a scene from known imagery. Recent approaches have afforded dramatic progress on small bounded regions of the scene. For an unbounded scene where cameras point in any direction and contents exist at any distance, certain mapping functions are used to represent it within a bounded space, yet they either work in object-centric scenes or focus on objects close to the camera. The goal of this paper is to understand how to design a proper mapping function that considers per-scene optimization, which remains unexplored. We first present a geometric understanding of existing mapping functions that express the relation between the bounded and unbounded scenes. Here, we exploit a stereographic projection method to explain failures of the mapping functions, where input ray samples are too sparse to account for scene geometry in unbounded regions. To overcome the failures, we propose a novel mapping function based on a $p$-norm distance, allowing to adaptively sample the rays by adjusting the $p$-value according to scene geometry, even in unbounded regions. To take the advantage of our mapping function, we also introduce a new ray parameterization to properly allocate ray samples in the geometry of unbounded regions. Through the incorporation of both the novel mapping function and the ray parameterization within existing NeRF frameworks, our method achieves state-of-the-art novel view synthesis results on a variety of challenging datasets.

Geometry-Aware Projective Mapping for Unbounded Neural Radiance Fields

https://openreview.net/forum?id=w7BwaDHppp

Spotlight Poster

Much of the knowledge encoded in transformer language models (LMs) may be expressed in terms of relations: relations between words and their synonyms, entities and their attributes, etc. We show that, for a subset of relations, this computation is well-approximated by a single linear transformation on the subject representation. Linear relation representations may be obtained by constructing a first-order approximation to the LM from a single prompt, and they exist for a variety of factual, commonsense, and linguistic relations. However, we also identify many cases in which LM predictions capture relational knowledge accurately, but this knowledge is not linearly encoded in their representations. Our results thus reveal a simple, interpretable, but heterogeneously deployed knowledge representation strategy in transformer LMs.

Linearity of Relation Decoding in Transformer Language Models

2308.09124

https://openreview.net/forum?id=w7LU2s14kE

Poster

We introduce a method to train vision-language models for remote-sensing images without using any textual annotations. Our key insight is to use co-located internet imagery taken on the ground as an intermediary for connecting remote-sensing images and language. Specifically, we train an image encoder for remote sensing images to align with the image encoder of CLIP using a large amount of paired internet and satellite images. Our unsupervised approach enables the training of a first-of-its-kind large scale VLM for remote sensing images at two different resolutions. We show that these VLMs enable zero-shot, open-vocabulary image classification, retrieval, segmentation and visual question answering for satellite images. On each of these tasks, our VLM trained without textual annotations outperforms existing VLMs trained with supervision, with gains of up to 20\% for classification and 80\% for segmentation.

Remote Sensing Vision-Language Foundation Models without Annotations via Ground Remote Alignment

2312.06960

https://openreview.net/forum?id=w9tc699w3Z

Poster

Constrained cooperative multi-agent reinforcement learning (MARL) is an emerging learning framework that has been widely applied to manage multi-agent systems, and many primal-dual type algorithms have been developed for it. However, the convergence of primal-dual algorithms crucially relies on strong duality -- a condition that has not been formally proved in constrained cooperative MARL. In this work, we prove that strong duality fails to hold in constrained cooperative MARL, by revealing a nonconvex quadratic type constraint on the occupation measure induced by the product policy. Consequently, our reanalysis of the primal-dual algorithm shows that its convergence rate is hindered by the nonzero duality gap. Then, we propose a decentralized primal approach for constrained cooperative MARL to avoid the duality gap, and our analysis shows that its convergence is hindered by another gap induced by the advantage functions. Moreover, we compare these two types of algorithms via concrete examples, and show that neither of them always outperforms the other one. Our study reveals that constrained cooperative MARL is generally a challenging and highly nonconvex problem, and its fundamental structure is very different from that of single-agent constrained RL.

On the Hardness of Constrained Cooperative Multi-Agent Reinforcement Learning

https://openreview.net/forum?id=wFWuX1Fhtj

Poster

An essential and challenging problem in causal inference is causal effect estimation from observational data. The problem becomes more difficult with the presence of unobserved confounding variables. The front-door adjustment is a practical approach for dealing with unobserved confounding variables. However, the restriction for the standard front-door adjustment is difficult to satisfy in practice. In this paper, we relax some of the restrictions by proposing the concept of conditional front-door (CFD) adjustment and develop the theorem that guarantees the causal effect identifiability of CFD adjustment. Furthermore, as it is often impossible for a CFD variable to be given in practice, it is desirable to learn it from data. By leveraging the ability of deep generative models, we propose CFDiVAE to learn the representation of the CFD adjustment variable directly from data with the identifiable Variational AutoEncoder and formally prove the model identifiability. Extensive experiments on synthetic datasets validate the effectiveness of CFDiVAE and its superiority over existing methods. The experiments also show that the performance of CFDiVAE is less sensitive to the causal strength of unobserved confounding variables. We further apply CFDiVAE to a real-world dataset to demonstrate its potential application.

Causal Inference with Conditional Front-Door Adjustment and Identifiable Variational Autoencoder

2310.01937

https://openreview.net/forum?id=wFf9m4v7oC

Poster

While score-based generative models (SGMs) have achieved remarkable successes in enormous image generation tasks, their mathematical foundations are still limited. In this paper, we analyze the approximation and generalization of SGMs in learning a family of sub-Gaussian probability distributions. We introduce a measure of complexity for probability distributions in terms of their relative density with respect to the standard Gaussian measure. We prove that if the log-relative density can be locally approximated by a neural network whose parameters can be suitably bounded, then the distribution generated by empirical score matching approximates the target distribution in total variation with a dimension-independent rate. We illustrate our theory through examples, which include certain mixtures of Gaussians. An essential ingredient of our proof is to derive a dimension-free deep network approximation rate for the true score function associated to the forward process, which is interesting in its own right.

Score-based generative models break the curse of dimensionality in learning a family of sub-Gaussian distributions

2402.08082

https://openreview.net/forum?id=wG12xUSqrI

Poster

Rotary Position Embeddings (RoPE) have been shown to effectively encode positional information in transformer-based language models. However, these models fail to generalize past the sequence length they were trained on. We present YaRN (Yet another RoPE extensioN method), a compute-efficient method to extend the context window of such models, requiring 10x less tokens and 2.5x less training steps than previous methods. Using YaRN, we show that LLaMA models can effectively utilize and extrapolate to context lengths much longer than their original pre-training would allow, while also surpassing previous the state-of-the-art at context window extension. In addition, we demonstrate that YaRN exhibits the capability to extrapolate beyond the limited context of a fine-tuning dataset. The models fine-tuned using YaRN has been made available and reproduced online up to 128k context length.

YaRN: Efficient Context Window Extension of Large Language Models

2309.00071

https://openreview.net/forum?id=wHBfxhZu1u

Spotlight Poster

Pre-training followed by full fine-tuning has gradually been substituted by Parameter-Efficient Tuning (PET) in the field of computer vision tasks. PET has gained popularity, especially in the context of large-scale models, due to its ability to reduce transfer learning costs and conserve hardware resources. However, existing PET approaches primarily focus on recognition tasks and typically support uni-modal optimization, neglecting dense prediction tasks and vision language interactions. To address this limitation, we propose a novel PET framework called Bi-directional Intertwined Vision Language Efficient Tuning for Referring Image Segmentation (BarLeRIa), which leverages bi-directional intertwined vision language adapters to fully exploit the frozen pre-trained models' potential in cross-modal dense prediction tasks. In BarLeRIa, two different tuning modules are employed for efficient global and local attention, as well as an intertwined vision language tuning algorithm for efficient modal fusion. Extensive experiments conducted on challenging RefCOCO-related benchmarks demonstrating the superiority of BarLeRIa over prior PET methods with a significant margin, \emph{i.e.}, achieving an average improvement of 5.6\%. Remarkably, without requiring additional training datasets, BarLeRIa even surpasses SOTA full fine-tuning approaches.

BarLeRIa: An Efficient Tuning Framework for Referring Image Segmentation

https://openreview.net/forum?id=wHLDHRkmEu

Spotlight Poster

Generalized Linear Models (GLMs) encompass a wide array of regression and classification models, where prediction is a function of a linear combination of the input variables. Often in real-world scenarios, a number of observations would be added into or removed from the existing training dataset, necessitating the development of learning systems that can efficiently train optimal models with varying observations in an online (sequential) manner instead of retraining from scratch. Despite the significance of data-varying scenarios, most existing approaches to sparse GLMs concentrate on offline batch updates, leaving online solutions largely underexplored. In this work, we present the first algorithm without compromising accuracy for GLMs regularized by sparsity-enforcing penalties trained on varying observations. Our methodology is capable of handling the addition and deletion of observations simultaneously, while adaptively updating data-dependent regularization parameters to ensure the best statistical performance. Specifically, we recast sparse GLMs as a bilevel optimization objective upon varying observations and characterize it as an explicit gradient flow in the underlying space for the inner and outer subproblems we are optimizing over, respectively. We further derive a set of rules to ensure a proper transition at regions of non-smoothness, and establish the guarantees of theoretical consistency and finite convergence. Encouraging results are exhibited on real-world benchmarks.

Learning No-Regret Sparse Generalized Linear Models with Varying Observation(s)

https://openreview.net/forum?id=wISvONp3Kq

Poster

In order to solve a task using reinforcement learning, it is necessary to first formalise the goal of that task as a *reward function*. However, for many real-world tasks, it is very difficult to manually specify a reward function that never incentivises undesirable behaviour. As a result, it is increasingly popular to use *reward learning algorithms*, which attempt to *learn* a reward function from data. However, the theoretical foundations of reward learning are not yet well-developed. In particular, it is typically not known when a given reward learning algorithm with high probability will learn a reward function that is safe to optimise. This means that reward learning algorithms generally must be evaluated empirically, which is expensive, and that their failure modes are difficult to predict in advance. One of the roadblocks to deriving better theoretical guarantees is the lack of good methods for *quantifying* the difference between reward functions. In this paper we provide a solution to this problem, in the form of a class of pseudometrics on the space of all reward functions that we call STARC (STAndardised Reward Comparison) metrics. We show that STARC metrics induce both an upper and a lower bound on worst-case regret, which implies that our metrics are tight, and that any metric with the same properties must be bilipschitz equivalent to ours. Moreover, we also identify a number of issues with reward metrics proposed by earlier works. Finally, we evaluate our metrics empirically, to demonstrate their practical efficacy. STARC metrics can be used to make both theoretical and empirical analysis of reward learning algorithms both easier and more principled.

STARC: A General Framework For Quantifying Differences Between Reward Functions

2309.15257

https://openreview.net/forum?id=wPhbtwlCDa

Poster

Visual prompting (VP) is an emerging parameter-efficient fine-tuning approach to adapting pre-trained vision models to solve various downstream image-classification tasks. However, there has hitherto been little systematic study of the design space of VP and no clear benchmark for evaluating its performance. To bridge this gap, we propose AutoVP, an end-to-end expandable framework for automating VP design choices, along with 12 downstream image-classification tasks that can serve as a holistic VP-performance benchmark. Our design space covers 1) the joint optimization of the prompts; 2) the selection of pre-trained models, including image classifiers and text-image encoders; and 3) model output mapping strategies, including nonparametric and trainable label mapping. Our extensive experimental results show that AutoVP outperforms the best-known current VP methods by a substantial margin, having up to 6.7% improvement in accuracy; and attains a maximum performance increase of 27.5% compared to linear-probing (LP) baseline. AutoVP thus makes a two-fold contribution: serving both as an efficient tool for hyperparameter tuning on VP design choices, and as a comprehensive benchmark that can reasonably be expected to accelerate VP’s development.

AutoVP: An Automated Visual Prompting Framework and Benchmark

2310.08381

https://openreview.net/forum?id=wR9qVlPh0P

Poster

Recent advancements in learning-based Multi-View Stereo (MVS) methods have prominently featured transformer-based models with attention mechanisms. However, existing approaches have not thoroughly investigated the profound influence of transformers on different MVS modules, resulting in limited depth estimation capabilities. In this paper, we introduce MVSFormer++, a method that prudently maximizes the inherent characteristics of attention to enhance various components of the MVS pipeline. Formally, our approach involves infusing cross-view information into the pre-trained DINOv2 model to facilitate MVS learning. Furthermore, we employ different attention mechanisms for the feature encoder and cost volume regularization, focusing on feature and spatial aggregations respectively. Additionally, we uncover that some design details would substantially impact the performance of transformer modules in MVS, including normalized 3D positional encoding, adaptive attention scaling, and the position of layer normalization. Comprehensive experiments on DTU, Tanks-and-Temples, BlendedMVS, and ETH3D validate the effectiveness of the proposed method. Notably, MVSFormer++ achieves state-of-the-art performance on the challenging DTU and Tanks-and-Temples benchmarks. Codes and models are available at https://github.com/maybeLx/MVSFormerPlusPlus.

MVSFormer++: Revealing the Devil in Transformer's Details for Multi-View Stereo

2401.11673

https://openreview.net/forum?id=wXWfvSpYHh

Poster

Transformer-based Large Language Models (LLMs) are pioneering advances in many natural language processing tasks, however, their exceptional capabilities are restricted within the preset context window of Transformer. Position Embedding (PE) scaling methods, while effective in extending the context window to a specific length, demonstrate either notable limitations in their extrapolation abilities or sacrificing partial performance within the context window. Length extrapolation methods, although theoretically capable of extending the context window beyond the training sequence length, often underperform in practical long-context applications. To address these challenges, we propose Continuous Length EXtrapolation (CLEX) for LLMs. We generalise the PE scaling approaches to model the continuous dynamics by ordinary differential equations over the length scaling factor, thereby overcoming the constraints of current PE scaling methods designed for specific lengths. Moreover, by extending the dynamics to desired context lengths beyond the training sequence length, CLEX facilitates the length extrapolation with impressive performance in practical tasks. We demonstrate that CLEX can be seamlessly incorporated into LLMs equipped with Rotary Position Embedding, such as LLaMA and GPT-NeoX, with negligible impact on training and inference latency. Experimental results reveal that CLEX can effectively extend the context window to over 4x training length, with no deterioration in performance. Furthermore, when evaluated on the practical LongBench benchmark, our model trained on a 4k length exhibits competitive performance against state-of-the-art open-source models trained on context lengths up to 32k.

CLEX: Continuous Length Extrapolation for Large Language Models

2310.16450

https://openreview.net/forum?id=wXpSidPpc5

Poster

In this work, we theoretically investigate the generalization property of neural networks (NN) trained by stochastic gradient descent (SGD) with \emph{large learning rate}. Under such a training regime, our finding is that, the oscillation of the NN weights caused by SGD with large learning rates turns out to be beneficial to generalization, potentially improving over the same NN trained by SGD with small learning rates that converges more smoothly. In view of the findings, we call such a phenomenon “benign oscillation”. Our theory towards demystifying such a phenomenon builds upon the feature learning perspective of deep learning. Specifically, we consider a feature-noise data generation model that consists of (i) weak features which have a small $\ell_2$-norm and appear in each data point; (ii) strong features which have a large $\ell_2$-norm but appear only in a certain fraction of all data points; and (iii) noise. We prove that NNs trained by oscillating SGD with a large learning rate can effectively learn the weak features in the presence of those strong features. In contrast, NNs trained by SGD with a small learning rate only learn the strong features but make little progress in learning the weak features. Consequently, when it comes to the new testing data points that consist of only weak features, the NN trained by oscillating SGD with large learning rates can still make correct predictions, while the NN trained by SGD with small learning rates could not. Our theory sheds light on how large learning rate training benefits the generalization of NNs. Experimental results demonstrate our findings on the phenomenon of “benign oscillation".

Benign Oscillation of Stochastic Gradient Descent with Large Learning Rate

2310.17074

https://openreview.net/forum?id=wYmvN3sQpG

Poster

Realistic graphs contain both rich self-features and informative neighborhood structures, jointly handled by a GNN in the typical setup. We propose to decouple the two modalities by mixture of weak and strong experts (Mowst), where the weak expert is a light-weight Multi-layer Perceptron (MLP) , and the strong expert is an off-the-shelf Graph Neural Network (GNN). To adapt the experts' collaboration to different target nodes, we propose a "confidence" mechanism based on the dispersion of the weak expert's prediction logits. The strong expert is conditionally activated in the low-confidence region when either the node's classification relies on neighborhood information, or the weak expert has low model quality. We reveal interesting training dynamics by analyzing the influence of the confidence function on loss: our training algorithm encourages specialization of each expert by effectively generating a soft splitting of the graph. In addition, our "confidence" design imposes a desirable bias towards the strong expert to benefit from the better generalization capability of GNNs. Mowst is easy to optimize and achieves strong expressive power, with computation cost comparable to a single GNN. Empirically, Mowst shows significant accuracy improvement on 6 standard node classification benchmarks (including both homophilous and heterophilous graphs).

Mixture of Weak and Strong Experts on Graphs

https://openreview.net/forum?id=wYvuY60SdD

Poster

Generating bitmap graphics from text has gained considerable attention, yet for scientific figures, vector graphics are often preferred. Given that vector graphics are typically encoded using low-level graphics primitives, generating them directly is difficult. To address this, we propose the use of TikZ, a well-known abstract graphics language that can be compiled to vector graphics, as an intermediate representation of scientific figures. TikZ offers human-oriented, high-level commands, thereby facilitating conditional language modeling with any large language model. To this end, we introduce DaTikZ the first large-scale TikZ dataset, consisting of 120k TikZ drawings aligned with captions. We fine-tune LLaMA on DaTikZ, as well as our new model CLiMA, which augments LLaMA with multimodal CLIP embeddings. In both human and automatic evaluation, CLiMA and LLaMA outperform commercial GPT-4 and Claude 2 in terms of similarity to human-created figures, with CLiMA additionally improving text-image alignment. Our detailed analysis shows that all models generalize well and are not susceptible to memorization. GPT-4 and Claude 2, however, tend to generate more simplistic figures compared to both humans and our models. We make our framework, AutomaTikZ, along with model weights and datasets, publicly available.

AutomaTikZ: Text-Guided Synthesis of Scientific Vector Graphics with TikZ

2310.00367

https://openreview.net/forum?id=v3K5TVP8kZ

Poster

Deploying reinforcement learning (RL) systems requires robustness to uncertainty and model misspecification, yet prior robust RL methods typically only study noise introduced independently across time. However, practical sources of uncertainty are usually coupled across time.We formally introduce temporally-coupled perturbations, presenting a novel challenge for existing robust RL methods. To tackle this challenge, we propose GRAD, a novel game-theoretic approach that treats the temporally-coupled robust RL problem as a partially-observable two-player zero-sum game. By finding an approximate equilibrium within this game, GRAD optimizes for general robustness against temporally-coupled perturbations. Experiments on continuous control tasks demonstrate that, compared with prior methods, our approach achieves a higher degree of robustness to various types of attacks on different attack domains, both in settings with temporally-coupled perturbations and decoupled perturbations.

Game-Theoretic Robust Reinforcement Learning Handles Temporally-Coupled Perturbations

2307.12062

https://openreview.net/forum?id=wZWTHU7AsQ

Poster

Causal effects are usually studied in terms of the means of counterfactual distributions, which may be insufficient in many scenarios. Given a class of densities known up to normalizing constants, we propose to model counterfactual distributions by minimizing kernel Stein discrepancies in a doubly robust manner. This enables the estimation of counterfactuals over large classes of distributions while exploiting the desired double robustness. We present a theoretical analysis of the proposed estimator, providing sufficient conditions for consistency and asymptotic normality, as well as an examination of its empirical performance.

Counterfactual Density Estimation using Kernel Stein Discrepancies

2309.16129

https://openreview.net/forum?id=wZXlEFO3tZ

Spotlight Poster

We introduce the FRactional-Order graph Neural Dynamical network (FROND), a learning framework that extends traditional graph neural ordinary differential equation (ODE) models by incorporating the time-fractional Caputo derivative. Due to its non-local nature, fractional calculus allows our framework to capture long-term memories in the feature updating process, in contrast to the Markovian nature of updates in traditional graph neural ODE models. This can lead to improved graph representation learning.We offer an interpretation of the feature updating process on graphs from a non-Markovian random walk perspective when the feature updating is governed by a diffusion process. We demonstrate analytically that over-smoothing can be mitigated in this setting.To experimentally demonstrate the versatility of the FROND framework, we evaluate the fractional counterparts of various established graph ODE models. Their consistently superior performance, compared to their original counterparts, highlights the potential of the FROND framework as an effective extension to boost the efficacy of various graph neural ODE models.

Unleashing the Potential of Fractional Calculus in Graph Neural Networks with FROND

2404.17099

https://openreview.net/forum?id=wcka3bd7P4

Poster

Deep neural networks and large-scale datasets have revolutionized the field of machine learning. However, these large networks are susceptible to overfitting to label noise, resulting in reduced generalization. To address this challenge, two promising approaches have emerged: i) loss reweighting, which reduces the influence of noisy examples on the training loss, and ii) label correction that replaces noisy labels with estimated true labels. These directions have been pursued separately or combined as independent methods, lacking a unified approach. In this work, we present a unified method that seamlessly combines loss reweighting and label correction to enhance robustness against label noise in classification tasks. Specifically, by leveraging ideas from compositional data analysis in statistics, we frame the problem as a regression task, where loss reweighting and label correction can naturally be achieved with a shifted Gaussian label noise model. Our unified approach achieves strong performance compared to recent baselines on several noisy labeled datasets. We believe this work is a promising step towards robust deep learning in the presence of label noise.

Robust Classification via Regression for Learning with Noisy Labels

https://openreview.net/forum?id=wfgZc3IMqo

Spotlight Poster

Scaling up representations for images or text has been extensively investigated in the past few years and has led to revolutions in learning vision and language. However, scalable representation for 3D objects and scenes is relatively unexplored. In this work, we present Uni3D, a 3D foundation model to explore the unified 3D representation at scale. Uni3D uses a 2D initialized ViT end-to-end pretrained to align the 3D point cloud features with the image-text aligned features. Via the simple architecture and pretext task, Uni3D can leverage abundant 2D pretrained models as initialization and image-text aligned models as the target, unlocking the great potential of 2D model zoos and scaling-up strategies to the 3D world. We efficiently scale up Uni3D to one billion parameters, and set new records on a broad range of 3D tasks, such as zero-shot classification, few-shot classification, open-world understanding and zero-shot part segmentation. We show that the strong Uni3D representation also enables applications such as 3D painting and retrieval in the wild. We believe that Uni3D provides a new direction for exploring both scaling up and efficiency of the representation in 3D domain.

Uni3D: Exploring Unified 3D Representation at Scale

2310.06773

https://openreview.net/forum?id=wcaE4Dfgt8

Poster

Text-to-image generative models have garnered immense attention for their ability to produce high-fidelity images from text prompts. Among these, Stable Diffusion distinguishes itself as a leading open-source model in this fast-growing field. However, the intricacies of fine-tuning these models pose multiple challenges from new methodology integration to systematic evaluation. Addressing these issues, this paper introduces LyCORIS (Lora beYond Conventional methods, Other Rank adaptation Implementations for Stable diffusion), an open-source library that offers a wide selection of fine-tuning methodologies for Stable Diffusion. Furthermore, we present a thorough framework for the systematic assessment of varied fine-tuning techniques. This framework employs a diverse suite of metrics and delves into multiple facets of fine-tuning, including hyperparameter adjustments and the evaluation with different prompt types across various concept categories. Through this comprehensive approach, our work provides essential insights into the nuanced effects of fine-tuning parameters, bridging the gap between state-of-the-art research and practical application.

Navigating Text-To-Image Customization: From LyCORIS Fine-Tuning to Model Evaluation

2309.14859

https://openreview.net/forum?id=wfzXa8e783

Spotlight Poster

Deep implicit functions (DIFs) have emerged as a powerful paradigm for many computer vision tasks such as 3D shape reconstruction, generation, registration, completion, editing, and understanding. However, given a set of 3D shapes with associated covariates there is at present no shape representation method which allows to precisely represent the shapes while capturing the individual dependencies on each covariate. Such a method would be of high utility to researchers to discover knowledge hidden in a population of shapes. For scientific shape discovery purpose, we propose a 3D Neural Additive Model for Interpretable Shape Representation ($\texttt{NAISR}$) which describes individual shapes by deforming a shape atlas in accordance to the effect of disentangled covariates. Our approach captures shape population trends and allows for patient-specific predictions through shape transfer. $\texttt{NAISR}$ is the first approach to combine the benefits of deep implicit shape representations with an atlas deforming according to specified covariates. We evaluate $\texttt{NAISR}$ with respect to shape reconstruction, shape disentanglement, shape evolution, and shape transfer on three datasets, i.e. 1) $\textit{Starman}$, a simulated 2D shape dataset; 2) ADNI hippocampus 3D shape dataset; 3) pediatric airway 3D shape dataset. Our experiments demonstrate that $\texttt{NAISR}$ achieves competitive shape reconstruction performance while retaining interpretability.

$\texttt{NAISR}$: A 3D Neural Additive Model for Interpretable Shape Representation

2303.09234

https://openreview.net/forum?id=wg8NPfeMF9

Poster

In a distributed machine learning setting like Federated Learning where there are multiple clients involved which update their individual weights to a single central server, often training on the entire individual client's dataset for each client becomes cumbersome. To address this issue we propose CORESET-PFEDBAYES: a personalized coreset weighted federated learning setup where the training updates for each individual clients are forwarded to the central server based on only individual client coreset based representative data points instead of the entire client data. Through theoretical analysis we present how the average generalization error is minimax optimal up to logarithm bounds $\mathcal{O}(n_k^{-\frac{2 \beta}{2 \beta+d}} \log ^{2 \delta^{\prime}}(n_k))$, where $n_k$ denotes the coreset size and how the approximation error on the data likelihood differs from a vanilla Federated Learning setup as a function $G(\boldsymbol{w})$ of the coreset weights $\boldsymbol{w}$. Our experiments on different benchmark datasets based on a variety of recent personalized federated learning architectures show significant gains (+4.87\% on MNIST, +8.61\% on FashionMNIST, +9.71\% on CIFAR in terms of model accuracy across ) as compared to random sampling on the training data followed by federated learning, thereby indicating how intelligently selecting such training samples can help in performance. Additionally, through experiments on medical datasets our proposed method showcases some gains (e.g. +9.74\% under COVID-19 dataset) as compared to other submodular optimization based approaches used for subset selection on client's data.

Bayesian Coreset Optimization for Personalized Federated Learning

https://openreview.net/forum?id=uz7d2N2zul

Poster

Bandwidth constraints during signal acquisition frequently impede real-time detection applications. Hyperspectral data is a notable example, whose vast volume compromises real-time hyperspectral detection. To tackle this hurdle, we introduce a novel approach leveraging pre-acquisition modulation to reduce the acquisition volume. This modulation process is governed by a deep learning model, utilizing prior information. Central to our approach is LUM-ViT, a Vision Transformer variant. Uniquely, LUM-ViT incorporates a learnable under-sampling mask tailored for pre-acquisition modulation. To further optimize for optical calculations, we propose a kernel-level weight binarization technique and a three-stage fine-tuning strategy. Our evaluations reveal that, by sampling a mere 10\% of the original image pixels, LUM-ViT maintains the accuracy loss within 1.8\% on the ImageNet classification task. The method sustains near-original accuracy when implemented on real-world optical hardware, demonstrating its practicality.

LUM-ViT: Learnable Under-sampling Mask Vision Transformer for Bandwidth Limited Optical Signal Acquisition

https://openreview.net/forum?id=wkbeqr5XhC

Poster

Generative models trained on internet-scale data are capable of generating novel and realistic texts, images, and videos. A natural next question is whether these models can advance science, for example by generating novel stable materials. Traditionally, models with explicit structures (e.g., graphs) have been used in modeling structural relationships in scientific data (e.g., atoms and bonds in crystals), but generating structures can be difficult to scale to large and complex systems. Another challenge in generating materials is the mismatch between standard generative modeling metrics and downstream applications. For instance, common metrics such as the reconstruction error do not correlate well with the downstream goal of discovering novel stable materials. In this work, we tackle the scalability challenge by developing a unified crystal representation that can represent any crystal structure (UniMat), followed by training a diffusion probabilistic model on these UniMat representations. Our empirical results suggest that despite the lack of explicit structure modeling, UniMat can generate high fidelity crystal structures from larger and more complex chemical systems, outperforming previous graph-based approaches under various generative modeling metrics. To better connect the generation quality of materials to downstream applications, such as discovering novel stable materials, we propose additional metrics for evaluating generative models of materials, including per-composition formation energy and stability with respect to convex hulls through decomposition energy from Density Function Theory (DFT). Lastly, we show that conditional generation with UniMat can scale to previously established crystal datasets with up to millions of crystals structures, outperforming random structure search (the current leading method for structure discovery) in discovering new stable materials.

Scalable Diffusion for Materials Generation

2311.09235

https://openreview.net/forum?id=wm4WlHoXpC

Spotlight Poster

Inverse design, where we seek to design input variables in order to optimize an underlying objective function, is an important problem that arises across fields such as mechanical engineering to aerospace engineering. Inverse design is typically formulated as an optimization problem, with recent works leveraging optimization across learned dynamics models. However, as models are optimized they tend to fall into adversarial modes, preventing effective sampling. We illustrate that by instead optimizing over the learned energy function captured by the diffusion model, we can avoid such adversarial examples and significantly improve design performance. We further illustrate how such a design system is compositional, enabling us to combine multiple different diffusion models representing subcomponents of our desired system to design systems with every specified component. In an N-body interaction task and a challenging 2D multi-airfoil design task, we demonstrate that by composing the learned diffusion model at test time, our method allows us to design initial states and boundary shapes that are more complex than those in the training data. Our method outperforms state-of-the-art neural inverse design method by an average of 41.5% in prediction MAE and 14.3% in design objective for the N-body dataset and discovers formation flying to minimize drag in the multi-airfoil design task.

Compositional Generative Inverse Design

2401.13171

https://openreview.net/forum?id=wmX0CqFSd7

Poster

Optimal Transport is a useful metric to compare probability distributions and to compute a pairing given a ground cost. Its entropic regularization variant (eOT) is crucial to have fast algorithms and reflect fuzzy/noisy matchings. This work focuses on Inverse Optimal Transport (iOT), the problem of inferring the ground cost from samples drawn from a coupling that solves an eOT problem. It is a relevant problem that can be used to infer unobserved/missing links, and to obtain meaningful information about the structure of the ground cost yielding the pairing. On one side, iOT benefits from convexity, but on the other side, being ill-posed, it requires regularization to handle the sampling noise. This work presents an in-depth theoretical study of the $\ell_1$ regularization to model for instance Euclidean costs with sparse interactions between features. Specifically, we derive a sufficient condition for the robust recovery of the sparsity of the ground cost that can be seen as a far reaching generalization of the Lasso’s celebrated ``Irrepresentability Condition’’. To provide additional insight into this condition (consequently on the types of recoverable costs) we work out in detail the Gaussian case. Surprisingly, varying the entropic regularizer provides evidence that the Gaussian iOT interpolates between a graphical Lasso and a classical Lasso, thereby establishing a connection between iOT and graph estimation, an important problem in ML.

Sparsistency for inverse optimal transport

2310.05461

https://openreview.net/forum?id=wpXGPCBOTX

Poster

The training method of Spiking Neural Networks (SNNs) is an essential problem, and how to integrate local and global learning is a worthy research interest. However, the current integration methods do not consider the network conditions suitable for local and global learning, and thus fail to balance their advantages. In this paper, we propose an Excitation-Inhibition Mechanism-assisted Hybrid Learning(EIHL) algorithm that adjusts the network connectivity by using the excitation-inhibition mechanism and then switches between local and global learning according to the network connectivity. The experimental results on CIFAR10/100 and DVS-CIFAR10 demonstrate that the EIHL not only has better accuracy performance than other methods but also has excellent sparsity advantage. Especially, the Spiking VGG11 is trained by EIHL, STBP, and STDP on DVS_CIFAR10, respectively. The accuracy of the Spiking VGG11 model on EIHL is 62.45%, which is 4.35% higher than STBP and 11.40% higher than STDP, and the sparsity is 18.74%, which is 18.74% higher than the other two methods. Moreover, the excitation-inhibition mechanism used in our method also offers a new perspective on the field of SNN learning.

Adaptive deep spiking neural network with global-local learning via balanced excitatory and inhibitory mechanism

https://openreview.net/forum?id=wpnlc2ONu0

Poster

We propose Consistency-guided Prompt learning (CoPrompt), a new fine-tuning method for vision-language models. Our approach improves the generalization of large foundation models when fine-tuned on downstream tasks in a few-shot setting. The basic idea of CoPrompt is to enforce a consistency constraint in the prediction of the trainable and pre-trained models to prevent overfitting on the downstream task. Additionally, we introduce the following two components into our consistency constraint to further boost the performance: enforcing consistency on two perturbed inputs and combining two dominant paradigms of tuning, prompting and adapter. Enforcing consistency on perturbed input serves to further regularize the consistency constraint, thereby improving generalization. Moreover, the integration of adapters and prompts not only enhances performance on downstream tasks but also offers increased tuning flexibility in both input and output spaces. This facilitates more effective adaptation to downstream tasks in a few-shot learning setting. Experiments show that CoPrompt outperforms existing methods on a range of evaluation suites, including base-to-novel generalization, domain generalization, and cross-dataset evaluation. On generalization, CoPrompt improves the state-of-the-art on zero-shot tasks and the overall harmonic mean over 11 datasets. Detailed ablation studies show the effectiveness of each of the components in CoPrompt. We make our code available at https://github.com/ShuvenduRoy/CoPrompt.

Consistency-guided Prompt Learning for Vision-Language Models

2306.01195

https://openreview.net/forum?id=wsRXwlwx4w

Poster

How to conduct teacher training for knowledge distillation is still an open problem. It has been widely observed that a best-performing teacher does not necessarily yield the best-performing student, suggesting a fundamental discrepancy between the current teacher training practice and the ideal teacher training strategy. To fill this gap, we explore the feasibility of training a teacher that is oriented toward student performance with empirical risk minimization (ERM). Our analyses are inspired by the recent findings that the effectiveness of knowledge distillation hinges on the teacher’s capability to approximate the true label distribution of training inputs. We theoretically establish that ERM minimizer can approximate the true label distribution of training data as long as the feature extractor of the learner network is Lipschitz continuous and is robust to feature transformations. In light of our theory, we propose a teacher training method SoTeacher which incorporates Lipschitz regularization and consistency regularization into ERM. Experiments on benchmark datasets using various knowledge distillation algorithms and teacher-student pairs confirm that SoTeacher can improve student accuracy consistently.

Toward Student-oriented Teacher Network Training for Knowledge Distillation

https://openreview.net/forum?id=wsWGcw6qKD

Poster

Large language models (LLMs) have shown significant improvements due to alignment tuning, that is, supervised fine-tuning (SFT) on instruction data and reinforcement learning from human feedback (RLHF).This raises questions about what is precisely learned during the alignment tuning process.We investigate the effects of alignment tuning through the lens of token distribution shift between untuned LLMs and their aligned counterparts (e.g., Llama-2 versus Llama-2-Chat).Our findings reveal that most distribution changes lie in stylistic tokens (e.g., transitional words, discourse markers), suggesting that LLMs primarily learn the language style of AI assistants during alignment tuning, while most of useful knowledge has been acquired by untuned LLMs. Thus, we pose the question: Is it necessary to update model weights to attain LLM alignment?Based on these insights, we propose an alternative method, Untuned LLMs with Restyled In-context Alignment (\textsc{Urial}), which achieves effective alignment solely through in-context learning (ICL) with as few as three curated, stylistic examples.Our evaluation on diverse examples from LIMA and AlpacaEval demonstrates that \textsc{Urial} can achieve highly satisfactory performance, sometimes equaling or surpassing SFT+RLHF counterparts, especially when the untuned LLM is sufficiently pre-trained.This implies that fine-tuning may not be as always crucial as previously assumed for LLM alignment, and lightweight alignment methods like \textsc{Urial} hold promise for efficiently tailoring LLM behavior without fine-tuning.

The Unlocking Spell on Base LLMs: Rethinking Alignment via In-Context Learning

https://openreview.net/forum?id=wxJ0eXwwda

Poster

We introduce ADoPD, a large-scale document page decomposition dataset for document understanding, encompassing document entity segmentation, text detection, tagging, and captioning. ADoPD stands out with its novel document taxonomy, meticulously crafted through a data-driven approach enriched by both large-scale pretrained models and human expertise. Our dataset achieves diversity by combining outlier detection with a human-in-the-loop approach. This significant contribution advances the field of document analysis, deepening our insights into document structures and substantially enhancing document processing and analysis techniques. The amalgamation of data-driven exploration, thorough annotation, and the human-in-the-loop methodology paves the way for innovative improvements in document analysis capabilities and the advancement of document processing applications. We conduct a comprehensive evaluation of ADoPD using various methods and demonstrate its effectiveness.

ADOPD: A Large-Scale Document Page Decomposition Dataset

https://openreview.net/forum?id=x1ptaXpOYa

Poster

In-context learning (ICL) is an important capability of Large Language Models (LLMs), enabling these models to dynamically adapt based on specific, in-context exemplars, thereby improving accuracy and relevance.However, LLM's responses may leak the sensitive private information contained in in-context exemplars. To address this challenge, we propose Differentially Private In-context Learning (DP-ICL), a general paradigm for privatizing ICL tasks. The key idea for DP-ICL paradigm is generating differentially private responses through a noisy consensus among an ensemble of LLM's responses based on disjoint exemplar sets. Based on the general paradigm of DP-ICL, we instantiate several techniques showing how to privatize ICL for text classification and language generation. We experiment on four text classification benchmarks and two language generation tasks, and our empirical findings suggest that our DP-ICL achieves a strong utility-privacy tradeoff.

Privacy-Preserving In-Context Learning for Large Language Models

2305.01639

https://openreview.net/forum?id=x4OPJ7lHVU

Poster

Test-time adaptation (TTA) aims to adapt a pre-trained model from a source domain to a target domain only using online unlabeled target data during testing, without accessing to the source data or modifying the original training process. Among the various TTA methods, pseudo-labeling has gained popularity. However, the presence of incorrect pseudo-labels can hinder the effectiveness of target domain adaptation. To overcome this challenge, we propose a novel TTA method, called PROtotype GRAph Model based pseudo-label learning (PROGRAM). PROGRAM consists of two key components: (1) Prototype Graph Model (PGM) for reliable pseudo-label generation; (2) Robust Self-Training (RST) for test-time adaptation with noisy pseudo-labels. PGM constructs the graph using prototypes and test samples, facilitating effective message passing among them to generate more reliable pseudo-labels. RST combines the advantages of consistency regularization and pseudo-labeling to achieve robust target domain adaptation in the presence of noisy pseudo-labels. Our proposed PROGRAM can be easily integrated into existing baselines, resulting in consistent improvement. Extensive experiments show that our PROGRAM outperforms the existing TTA methods on multiple domain generalization and image corruption benchmarks.

PROGRAM: PROtotype GRAph Model based Pseudo-Label Learning for Test-Time Adaptation

https://openreview.net/forum?id=x5LvBK43wg

Spotlight Poster

A growing literature in computational neuroscience leverages gradient descent and learning algorithms that approximate it to study synaptic plasticity in the brain. However, the vast majority of this work ignores a critical underlying assumption: the choice of distance for synaptic changes (i.e. the geometry of synaptic plasticity). Gradient descent assumes that the distance is Euclidean, but many other distances are possible, and there is no reason that biology necessarily uses Euclidean geometry. Here, using the theoretical tools provided by mirror descent, we show that, regardless of the loss being minimized, the distribution of synaptic weights will depend on the geometry of synaptic plasticity. We use these results to show that experimentally-observed log-normal weight distributions found in several brain areas are not consistent with standard gradient descent (i.e. a Euclidean geometry), but rather with non-Euclidean distances. Finally, we show that it should be possible to experimentally test for different synaptic geometries by comparing synaptic weight distributions before and after learning. Overall, this work shows that the current paradigm in theoretical work on synaptic plasticity that assumes Euclidean synaptic geometry may be misguided and that it should be possible to experimentally determine the true geometry of synaptic plasticity in the brain.

Synaptic Weight Distributions Depend on the Geometry of Plasticity

2305.19394

https://openreview.net/forum?id=x5txICnnjC

Poster

This paper presents Tag2Text, a vision language pre-training (VLP) framework, which introduces image tagging into vision-language models to guide the learning of visual-linguistic features. In contrast to prior works which utilize object tags either manually labeled or automatically detected with a limited detector, our approach utilizes tags parsed from its paired text to learn an image tagger and meanwhile provides guidance to vision-language models. Given that, Tag2Text can utilize large-scale annotation-free image tags in accordance with image-text pairs, and provides more diverse tag categories beyond objects. Strikingly, Tag2Text showcases the ability of a foundational image tagging model, with superior zero-shot performance even comparable to full supervision manner. Moreover, by leveraging tagging guidance, Tag2Text effectively enhances the performance of vision-language models on both generation-based and alignment-based tasks. Across a wide range of downstream benchmarks, Tag2Text achieves state-of-the-art results with similar model sizes and data scales, demonstrating the efficacy of the proposed tagging guidance.

Tag2Text: Guiding Vision-Language Model via Image Tagging

2303.05657

https://openreview.net/forum?id=x6u2BQ7xcq

Poster

Image denoisers have been shown to be powerful priors for solving inverse problems in imaging. In this work, we introduce a generalization of these methods that allows any image restoration network to be used as an implicit prior. The proposed method uses priors specified by deep neural networks pre-trained as general restoration operators. The method provides a principled approach for adapting state-of-the-art restoration models for other inverse problems. Our theoretical result analyzes its convergence to a stationary point of a global functional associated with the restoration operator. Numerical results show that the method using a super-resolution prior achieves state-of-the-art performance both quantitatively and qualitatively. Overall, this work offers a step forward for solving inverse problems by enabling the use of powerful pre-trained restoration models as priors.

A Restoration Network as an Implicit Prior

2310.01391

https://openreview.net/forum?id=x7d1qXEn1e

Poster

Designing effective positional encodings for graphs is key to building powerful graph transformers and enhancing message-passing graph neural networks. Although widespread, using Laplacian eigenvectors as positional encodings faces two fundamental challenges: (1) *Non-uniqueness*: there are many different eigendecompositions of the same Laplacian, and (2) *Instability*: small perturbations to the Laplacian could result in completely different eigenspaces, leading to unpredictable changes in positional encoding. Despite many attempts to address non-uniqueness, most methods overlook stability, leading to poor generalization on unseen graph structures. We identify the cause of instability to be the use of "hard partition'' of eigenspaces. Hence, we introduce Stable and Expressive Positional Encodings (SPE), an architecture for processing eigenvectors that uses eigenvalues to ``softly partition'' eigenspaces. SPE is the first architecture that is (1) provably stable, and (2) universally expressive for basis invariant functions whilst respecting all symmetries of eigenvectors. Besides guaranteed stability, we prove that SPE is at least as expressive as existing methods, and highly capable of counting graph structures. Finally, we evaluate the effectiveness of our method on molecular property prediction, and out-of-distribution generalization tasks, finding improved generalization compared to existing positional encoding methods.

On the Stability of Expressive Positional Encodings for Graphs

2310.02579

https://openreview.net/forum?id=xAqcJ9XoTf

Poster

Diffusion models have attracted significant attention due to the remarkable ability to create content and generate data for tasks like image classification. However, the usage of diffusion models to generate the high-quality object detection data remains an underexplored area, where not only image-level perceptual quality but also geometric conditions such as bounding boxes and camera views are essential. Previous studies have utilized either copy-paste synthesis or layout-to-image (L2I) generation with specifically designed modules to encode the semantic layouts. In this paper, we propose the GeoDiffusion, a simple framework that can flexibly translate various geometric conditions into text prompts and empower pre-trained text-to-image (T2I) diffusion models for high-quality detection data generation. Unlike previous L2I methods, our GeoDiffusion is able to encode not only the bounding boxes but also extra geometric conditions such as camera views in self-driving scenes. Extensive experiments demonstrate GeoDiffusion outperforms previous L2I methods while maintaining 4x training time faster. To the best of our knowledge, this is the first work to adopt diffusion models for layout-to-image generation with geometric conditions and demonstrate that L2I-generated images can be beneficial for improving the performance of object detectors.

GeoDiffusion: Text-Prompted Geometric Control for Object Detection Data Generation

2306.04607

https://openreview.net/forum?id=xBfQZWeDRH

Poster

Advances in machine learning are closely tied to the creation of datasets. While data documentation is widely recognized as essential to the reliability, reproducibility, and transparency of ML, we lack a systematic empirical understanding of current dataset documentation practices. To shed light on this question, here we take Hugging Face - one of the largest platforms for sharing and collaborating on ML models and datasets - as a prominent case study. By analyzing all 7,433 dataset documentation on Hugging Face, our investigation provides an overview of the Hugging Face dataset ecosystem and insights into dataset documentation practices, yielding 5 main findings: (1) The dataset card completion rate shows marked heterogeneity correlated with dataset popularity: While 86.0\% of the top 100 downloaded dataset cards fill out all sections suggested by Hugging Face community, only 7.9\% of dataset cards with no downloads complete all these sections. (2) A granular examination of each section within the dataset card reveals that the practitioners seem to prioritize Dataset Description and Dataset Structure sections, accounting for 36.2\% and 33.6\% of the total card length, respectively, for the most downloaded datasets. In contrast, the Considerations for Using the Data section receives the lowest proportion of content, accounting for just 2.1\% of the text. (3) By analyzing the subsections within each section and utilizing topic modeling to identify key topics, we uncover what is discussed in each section, and underscore significant themes encompassing both technical and social impacts, as well as limitations within the Considerations for Using the Data section. (4) Our findings also highlight the need for improved accessibility and reproducibility of datasets in the Usage sections. (5) In addition, our human annotation evaluation emphasizes the pivotal role of comprehensive dataset content in shaping individuals' perceptions of a dataset card's overall quality. Overall, our study offers a unique perspective on analyzing dataset documentation through large-scale data science analysis and underlines the need for more thorough dataset documentation in machine learning research.

Navigating Dataset Documentations in AI: A Large-Scale Analysis of Dataset Cards on HuggingFace

https://openreview.net/forum?id=xC8xh2RSs2

Poster

Recent developments in offline reinforcement learning have uncovered the immense potential of diffusion modeling, which excels at representing heterogeneous behavior policies. However, sampling from diffusion policies is considerably slow because it necessitates tens to hundreds of iterative inference steps for one action. To address this issue, we propose to extract an efficient deterministic inference policy from critic models and pretrained diffusion behavior models, leveraging the latter to directly regularize the policy gradient with the behaviordistribution’s score function during optimization. Our method enjoys powerful generative capabilities of diffusion modeling while completely circumventing the computationally intensive and time-consuming diffusion sampling scheme, both during training and evaluation. Extensive results on D4RL tasks show that our method boosts action sampling speed by more than 25 times compared with various leading diffusion-based methods in locomotion tasks, while still maintaining state-of-the-art performance.

Score Regularized Policy Optimization through Diffusion Behavior

2310.07297

https://openreview.net/forum?id=xCRr9DrolJ

Poster

Diffusion models have demonstrated impressive generative capabilities, but their exposure bias problem, described as the input mismatch between training and sampling, lacks in-depth exploration. In this paper, we systematically investigate the exposure bias problem in diffusion models by first analytically modelling the sampling distribution, based on which we then attribute the prediction error at each sampling step as the root cause of the exposure bias issue. Furthermore, we discuss potential solutions to this issue and propose an intuitive metric for it. Along with the elucidation of exposure bias, we propose a simple, yet effective, training-free method called Epsilon Scaling to alleviate the exposure bias. We show that Epsilon Scaling explicitly moves the sampling trajectory closer to the vector field learned in the training phase by scaling down the network output (Epsilon), mitigating the input mismatch between training and sampling. Experiments on various diffusion frameworks (ADM, DDPM/DDIM, EDM, LDM), unconditional and conditional settings, and deterministic vs. stochastic sampling verify the effectiveness of our method. Remarkably, our ADM-ES, as a SOTA stochastic sampler, obtains 2.17 FID on CIFAR-10 under 100-step unconditional generation.

Elucidating the Exposure Bias in Diffusion Models

2308.15321

https://openreview.net/forum?id=xEJMoj1SpX

Spotlight Poster

This paper rigorously shows how over-parameterization dramatically changes the convergence behaviors of gradient descent (GD) for the matrix sensing problem, where the goal is to recover an unknown low-rank ground-truth matrix from near-isotropic linear measurements.First, we consider the symmetric setting with the symmetric parameterization where $M^* \in \mathbb{R}^{n \times n}$ is a positive semi-definite unknown matrix of rank $r \ll n$, and one uses a symmetric parameterization $XX^\top$ to learn $M^*$. Here $X \in \mathbb{R}^{n \times k}$ with $k > r$ is the factor matrix. We give a novel $\Omega\left(1/T^2\right)$ lower bound of randomly initialized GD for the over-parameterized case ($k >r$) where $T$ is the number of iterations. This is in stark contrast to the exact-parameterization scenario ($k=r$) where the convergence rate is $\exp\left(-\Omega\left(T\right)\right)$. Next, we study asymmetric setting where $M^* \in \mathbb{R}^{n_1 \times n_2}$ is the unknown matrix of rank $r \ll \min\{n_1,n_2\}$, and one uses an asymmetric parameterization $FG^\top$ to learn $M^*$ where $F \in \mathbb{R}^{n_1 \times k}$ and $G \in \mathbb{R}^{n_2 \times k}$. We give the first global exact convergence result of randomly initialized GD for the exact-parameterization case ($k=r$) with an $\exp\left(-\Omega\left(T\right)\right)$ rate. Furthermore, we give the first global exact convergence result for the over-parameterization case ($k>r$) with an $\exp\left(-\Omega\left(\alpha^2 T\right)\right)$ rate where $\alpha$ is the initialization scale. This linear convergence result in the over-parameterization case is especially significant because one can apply the asymmetric parameterization to the symmetric setting to speed up from $\Omega\left(1/T^2\right)$ to linear convergence. Therefore, we identify a surprising phenomenon: asymmetric parameterization can exponentially speed up convergence. Equally surprising is our analysis that highlights the importance of imbalance between $F$ and $G$. This is in sharp contrast to prior works which emphasize balance. We further give an example showing the dependency on $\alpha$ in the convergence rate is unavoidable in the worst case. On the other hand, we propose a novel method that only modifies one step of GD and obtains a convergence rate independent of $\alpha$, recovering the rate in the exact-parameterization case. We provide empirical studies to verify our theoretical findings.

How Over-Parameterization Slows Down Gradient Descent in Matrix Sensing: The Curses of Symmetry and Initialization

2310.01769

https://openreview.net/forum?id=xGvPKAiOhq

Poster

Differentially private learning algorithms inject noise into the learning process. While the most common private learning algorithm, DP-SGD, adds independent Gaussian noise in each iteration, recent work on matrix factorization mechanisms has shown empirically that introducing correlations in the noise can greatly improve their utility. We characterize the asymptotic learning utility for any choice of the correlation function, giving precise analytical bounds for linear regression and as the solution to a convex program for general convex functions. We show, using these bounds, how correlated noise provably improves upon vanilla DP-SGD as a function of problem parameters such as the effective dimension and condition number. Moreover, our analytical expression for the near-optimal correlation function circumvents the cubic complexity of the semi-definite program used to optimize the noise correlation matrix in previous work. We validate these theoretical results with experiments on private deep learning. Our work matches or outperforms prior work while being efficient both in terms of computation and memory.

Correlated Noise Provably Beats Independent Noise for Differentially Private Learning

2310.06771

https://openreview.net/forum?id=xHmCdSArUC

Poster

Language Models (LMs) have greatly influenced diverse domains. However, their inherent limitation in comprehending 3D molecular structures has considerably constrained their potential in the biomolecular domain. To bridge this gap, we focus on 3D molecule-text interpretation, and propose 3D-MoLM: 3D-Molecular Language Modeling. Specifically, 3D-MoLM enables an LM to interpret andanalyze 3D molecules by equipping the LM with a 3D molecular encoder. This integration is achieved by a 3D molecule-text projector, bridging the 3D molecular encoder’s representation space and the LM’s input space. Moreover, to enhance 3DMoLM’s ability of cross-modal molecular understanding and instruction following, we meticulously curated a 3D molecule-centric instruction tuning dataset – 3D-MoIT. Through 3D molecule-text alignment and 3D molecule-centric instruction tuning, 3D-MoLM establishes an integration of 3D molecular encoder and LM. It significantly surpasses existing baselines on downstream tasks, including moleculetext retrieval, molecule captioning, and more challenging open-text molecular QA tasks, especially focusing on 3D-dependent properties. We will release our codes and datasets at https://anonymous.4open.science/r/3D-MoLM.

Towards 3D Molecule-Text Interpretation in Language Models

2401.13923

https://openreview.net/forum?id=xI4yNlkaqh

Spotlight Poster

Mixture models are traditionally represented and learned by adding several distributions as components. Allowing mixtures to subtract probability mass or density can drastically reduce the number of components needed to model complex distributions. However, learning such subtractive mixtures while ensuring they still encode a non-negative function is challenging. We investigate how to learn and perform inference on deep subtractive mixtures by squaring them. We do this in the framework of probabilistic circuits, which enable us to represent tensorized mixtures and generalize several other subtractive models. We theoretically prove that the class of squared circuits allowing subtractions can be exponentially more expressive than traditional additive mixtures; and, we empirically show this increased expressiveness on a series of real-world distribution estimation tasks.

Subtractive Mixture Models via Squaring: Representation and Learning

2310.00724

https://openreview.net/forum?id=xIHi5nxu9P

Spotlight Poster

This paper presents a new approach and algorithm for solving a class of constrained Bi-Level Optimization (BLO) problems in which the lower-level problem involves constraints coupling both upper-level and lower-level variables. Such problems have recently gained significant attention due to their broad applicability in machine learning. However, conventional gradient-based methods unavoidably rely on computationally intensive calculations related to the Hessian matrix. To address this challenge, we begin by devising a smooth proximal Lagrangian value function to handle the constrained lower-level problem. Utilizing this construct, we introduce a single-level reformulation for constrained BLOs that transforms the original BLO problem into an equivalent optimization problem with smooth constraints. Enabled by this reformulation, we develop a Hessian-free gradient-based algorithm—termed proximal Lagrangian Value function-based Hessian-free Bi-level Algorithm (LV-HBA)—that is straightforward to implement in a single loop manner. Consequently, LV-HBA is especially well-suited for machine learning applications. Furthermore, we offer non-asymptotic convergence analysis for LV-HBA, eliminating the need for traditional strong convexity assumptions for the lower-level problem while also being capable of accommodating non-singleton scenarios. Empirical results substantiate the algorithm's superior practical performance.

Constrained Bi-Level Optimization: Proximal Lagrangian Value Function Approach and Hessian-free Algorithm

2401.16164

https://openreview.net/forum?id=xJ5N8qrEPl

Spotlight Poster

Integral reinforcement learning (IntRL) demands the precise computation of the utility function's integral at its policy evaluation (PEV) stage. This is achieved through quadrature rules, which are weighted sums of utility functions evaluated from state samples obtained in discrete time. Our research reveals a critical yet underexplored phenomenon: the choice of the computational method -- in this case, the quadrature rule -- can significantly impact control performance. This impact is traced back to the fact that computational errors introduced in the PEV stage can affect the policy iteration's convergence behavior, which in turn affects the learned controller. To elucidate how computation impacts control, we draw a parallel between IntRL's policy iteration and Newton's method applied to the Hamilton-Jacobi-Bellman equation. In this light, computational error in PEV manifests as an extra error term in each iteration of Newton's method, with its upper bound proportional to the computational error. Further, we demonstrate that when the utility function resides in a reproducing kernel Hilbert space (RKHS), the optimal quadrature is achievable by employing Bayesian quadrature with the RKHS-inducing kernel function. We prove that the local convergence rates for IntRL using the trapezoidal rule and Bayesian quadrature with a Matérn kernel to be $O(N^{-2})$ and $O(N^{-b})$, where $N$ is the number of evenly-spaced samples and $b$ is the Matérn kernel's smoothness parameter. These theoretical findings are finally validated by two canonical control tasks.

Impact of Computation in Integral Reinforcement Learning for Continuous-Time Control

2402.17375

https://openreview.net/forum?id=xJEd8PkdNz

Spotlight Poster

Out-of-distribution (OOD) detection aims at identifying samples from unknown classes, playing a crucial role in trustworthy models against errors on unexpected inputs. There is extensive research dedicated to exploring OOD detection in the vision modality. Vision-language models (VLMs) can leverage both textual and visual information for various multi-modal applications, whereas few OOD detection methods take into account information from the text modality. In this paper, we propose a novel post hoc OOD detection method, called NegLabel, which takes a vast number of negative labels from extensive corpus databases. We design a novel scheme for the OOD score collaborated with negative labels.Theoretical analysis helps to understand the mechanism of negative labels. Extensive experiments demonstrate that our method NegLabel achieves state-of-the-art performance on various OOD detection benchmarks and generalizes well on multiple VLM architectures. Furthermore, our method NegLabel exhibits remarkable robustness against diverse domain shifts.

Negative Label Guided OOD Detection with Pretrained Vision-Language Models

2403.20078

https://openreview.net/forum?id=xUO1HXz4an

Spotlight Poster

Privacy amplification exploits randomness in data selection to provide tighter differential privacy (DP) guarantees. This analysis is key to DP-SGD's success in machine learning (ML), but, is not readily applicable to the newer state-of-the-art (SOTA) algorithms. This is because these algorithms, known as DP-FTRL, use the matrix mechanism to add correlated noise instead of independent noise as in DP-SGD.In this paper, we propose "MMCC'' (matrix mechanism conditional composition), the first algorithm to analyze privacy amplification via sampling for any generic matrix mechanism. MMCC is nearly tight in that it approaches a lower bound as $\epsilon\to0$. To analyze correlated outputs in MMCC, we prove that they can be analyzed as if they were independent, by conditioning them on prior outputs. Our "conditional composition theorem'' has broad utility: we use it to show that the noise added to binary-tree-DP-FTRL can asymptotically match the noise added to DP-SGD with amplification. Our algorithm also has practical empirical utility. We show that amplification leads to significant improvement in the privacy/utility trade-offs for DP-FTRL style algorithms for standard benchmark tasks.

Privacy Amplification for Matrix Mechanisms

2310.15526

https://openreview.net/forum?id=xUzWmFdglP

Poster

In human-written articles, we often leverage the subtleties of text style, such as bold and italics, to guide the attention of readers. These textual emphases are vital for the readers to grasp the conveyed information. When interacting with large language models (LLMs), we have a similar need -- steering the model to pay closer attention to user-specified information, e.g., an instruction. Existing methods, however, are constrained to process plain text and do not support such a mechanism. This motivates us to introduce PASTA -- Post-hoc Attention STeering Approach, a method that allows LLMs to read text with user-specified emphasis marks. To this end, PASTA identifies a small subset of attention heads and applies precise attention reweighting on them, directing the model attention to user-specified parts. Like prompting, PASTA is applied at inference time and does not require changing any model parameters. Experiments demonstrate that PASTA can substantially enhance an LLM's ability to follow user instructions or integrate new knowledge from user inputs, leading to a significant performance improvement on a variety of tasks, e.g., an average accuracy improvement of 22\% for LLAMA-7B. Code is provided at https://anonymous.4open.science/r/PASTA-10E9.

Tell Your Model Where to Attend: Post-hoc Attention Steering for LLMs

2311.02262

https://openreview.net/forum?id=xZDWO0oejD