ACL
Collection
Accepted papers for ACL (Annual Meeting of the Association for Computational Linguistics), one dataset per year. • 14 items • Updated
paper_id stringlengths 15 18 | title stringlengths 12 141 | paper_url stringlengths 41 44 | authors listlengths 1 44 | abstract large_stringlengths 410 1.78k | anthology_id stringlengths 15 18 | doi stringlengths 27 30 | award stringclasses 18
values | arxiv_id stringlengths 10 10 ⌀ | arxiv_id_source stringclasses 2
values |
|---|---|---|---|---|---|---|---|---|---|
2023.acl-long.1 | One Cannot Stand for Everyone! Leveraging Multiple User Simulators to train Task-oriented Dialogue Systems | https://aclanthology.org/2023.acl-long.1/ | [
"Yajiao Liu",
"Xin Jiang",
"Yichun Yin",
"Yasheng Wang",
"Fei Mi",
"Qun Liu",
"Xiang Wan",
"Benyou Wang"
] | User simulators are agents designed to imitate human users; recent advances have found that Task-oriented Dialogue (ToD) systems optimized toward a user simulator could better satisfy the need of human users. However, this might result in a sub-optimal ToD system if it is tailored to only one ad hoc user simulator, sin... | 2023.acl-long.1 | 10.18653/v1/2023.acl-long.1 | null | null | null |
2023.acl-long.2 | SafeConv: Explaining and Correcting Conversational Unsafe Behavior | https://aclanthology.org/2023.acl-long.2/ | [
"Mian Zhang",
"Lifeng Jin",
"Linfeng Song",
"Haitao Mi",
"Wenliang Chen",
"Dong Yu"
] | One of the main challenges open-domain end-to-end dialogue systems, or chatbots, face is the prevalence of unsafe behavior, such as toxic languages and harmful suggestions. However, existing dialogue datasets do not provide enough annotation to explain and correct such unsafe behavior. In this work, we construct a new ... | 2023.acl-long.2 | 10.18653/v1/2023.acl-long.2 | null | null | null |
2023.acl-long.3 | Detecting and Mitigating Hallucinations in Machine Translation: Model Internal Workings Alone Do Well, Sentence Similarity Even Better | https://aclanthology.org/2023.acl-long.3/ | [
"David Dale",
"Elena Voita",
"Loic Barrault",
"Marta R. Costa-jussà"
] | While the problem of hallucinations in neural machine translation has long been recognized, so far the progress on its alleviation is very little. Indeed, recently it turned out that without artificially encouraging models to hallucinate, previously existing methods fall short and even the standard sequence log-probabi... | 2023.acl-long.3 | 10.18653/v1/2023.acl-long.3 | null | 2212.08597 | title_snapshot |
2023.acl-long.4 | Explainable Recommendation with Personalized Review Retrieval and Aspect Learning | https://aclanthology.org/2023.acl-long.4/ | [
"Hao Cheng",
"Shuo Wang",
"Wensheng Lu",
"Wei Zhang",
"Mingyang Zhou",
"Kezhong Lu",
"Hao Liao"
] | Explainable recommendation is a technique that combines prediction and generation tasks to produce more persuasive results. Among these tasks, textual generation demands large amounts of data to achieve satisfactory accuracy. However, historical user reviews of items are often insufficient, making it challenging to ens... | 2023.acl-long.4 | 10.18653/v1/2023.acl-long.4 | null | 2306.12657 | title_snapshot |
2023.acl-long.5 | Binary and Ternary Natural Language Generation | https://aclanthology.org/2023.acl-long.5/ | [
"Zechun Liu",
"Barlas Oguz",
"Aasish Pappu",
"Yangyang Shi",
"Raghuraman Krishnamoorthi"
] | Ternary and binary neural networks enable multiplication-free computation and promise multiple orders of magnitude efficiency gains over full-precision networks if implemented on specialized hardware. However, since both the parameter and the output space are highly discretized, such networks have proven very difficult... | 2023.acl-long.5 | 10.18653/v1/2023.acl-long.5 | null | 2306.01841 | title_snapshot |
2023.acl-long.6 | Span-Selective Linear Attention Transformers for Effective and Robust Schema-Guided Dialogue State Tracking | https://aclanthology.org/2023.acl-long.6/ | [
"Björn Bebensee",
"Haejun Lee"
] | In schema-guided dialogue state tracking models estimate the current state of a conversation using natural language descriptions of the service schema for generalization to unseen services. Prior generative approaches which decode slot values sequentially do not generalize well to variations in schema, while discrimina... | 2023.acl-long.6 | 10.18653/v1/2023.acl-long.6 | null | 2306.09340 | title_snapshot |
2023.acl-long.7 | EM Pre-training for Multi-party Dialogue Response Generation | https://aclanthology.org/2023.acl-long.7/ | [
"Yiyang Li",
"Hai Zhao"
] | Dialogue response generation requires an agent to generate a response according to the current dialogue history, in terms of which two-party dialogues have been well studied, but leaving a great gap for multi-party dialogues at the same time. Different from two-party dialogues where each response is a direct reply to i... | 2023.acl-long.7 | 10.18653/v1/2023.acl-long.7 | null | 2305.12412 | title_snapshot |
2023.acl-long.8 | ACLM: A Selective-Denoising based Generative Data Augmentation Approach for Low-Resource Complex NER | https://aclanthology.org/2023.acl-long.8/ | [
"Sreyan Ghosh",
"Utkarsh Tyagi",
"Manan Suri",
"Sonal Kumar",
"Ramaneswaran S",
"Dinesh Manocha"
] | Complex Named Entity Recognition (NER) is the task of detecting linguistically complex named entities in low-context text. In this paper, we present ACLM Attention-map aware keyword selection for Conditional Language Model fine-tuning), a novel data augmentation approach based on conditional generation, to address the ... | 2023.acl-long.8 | 10.18653/v1/2023.acl-long.8 | null | 2306.00928 | title_snapshot |
2023.acl-long.9 | Natural Language to Code Generation in Interactive Data Science Notebooks | https://aclanthology.org/2023.acl-long.9/ | [
"Pengcheng Yin",
"Wen-Ding Li",
"Kefan Xiao",
"Abhishek Rao",
"Yeming Wen",
"Kensen Shi",
"Joshua Howland",
"Paige Bailey",
"Michele Catasta",
"Henryk Michalewski",
"Oleksandr Polozov",
"Charles Sutton"
] | Computational notebooks, such as Jupyter notebooks, are interactive computing environments that are ubiquitous among data scientists to perform data wrangling and analytic tasks. To measure the performance of AI pair programmers that automatically synthesize programs for those tasks given natural language (NL) intents ... | 2023.acl-long.9 | 10.18653/v1/2023.acl-long.9 | null | 2212.09248 | title_snapshot |
2023.acl-long.10 | Subset Retrieval Nearest Neighbor Machine Translation | https://aclanthology.org/2023.acl-long.10/ | [
"Hiroyuki Deguchi",
"Taro Watanabe",
"Yusuke Matsui",
"Masao Utiyama",
"Hideki Tanaka",
"Eiichiro Sumita"
] | k-nearest-neighbor machine translation (kNN-MT) (Khandelwal et al., 2021) boosts the translation performance of trained neural machine translation (NMT) models by incorporating example-search into the decoding algorithm. However, decoding is seriously time-consuming, i.e., roughly 100 to 1,000 times slower than standar... | 2023.acl-long.10 | 10.18653/v1/2023.acl-long.10 | null | null | null |
2023.acl-long.11 | MIL-Decoding: Detoxifying Language Models at Token-Level via Multiple Instance Learning | https://aclanthology.org/2023.acl-long.11/ | [
"Xu Zhang",
"Xiaojun Wan"
] | Despite advances in large pre-trained neural language models, they are prone to generating toxic language, which brings security risks to their applications. We introduce MIL-Decoding, which detoxifies language models at token-level by interpolating it with a trained multiple instance learning (MIL) network.MIL model i... | 2023.acl-long.11 | 10.18653/v1/2023.acl-long.11 | null | null | null |
2023.acl-long.12 | Dependency resolution at the syntax-semantics interface: psycholinguistic and computational insights on control dependencies | https://aclanthology.org/2023.acl-long.12/ | [
"Iria de-Dios-Flores",
"Juan Garcia Amboage",
"Marcos Garcia"
] | Using psycholinguistic and computational experiments we compare the ability of humans and several pre-trained masked language models to correctly identify control dependencies in Spanish sentences such as ‘José le prometió/ordenó a María ser ordenado/a’ (‘Joseph promised/ordered Mary to be tidy’). These structures unde... | 2023.acl-long.12 | 10.18653/v1/2023.acl-long.12 | null | null | null |
2023.acl-long.13 | Open-ended Long Text Generation via Masked Language Modeling | https://aclanthology.org/2023.acl-long.13/ | [
"Xiaobo Liang",
"Zecheng Tang",
"Juntao Li",
"Min Zhang"
] | Pre-trained autoregressive (AR) language models such as BART and GPTs have dominated OPen-ended Long Text Generation (Open-LTG).However, the AR nature will decrease the inference efficiency along with the increase of generation length, which hinder their application in Open-LTG.To improve inference efficiency, we alter... | 2023.acl-long.13 | 10.18653/v1/2023.acl-long.13 | null | null | null |
2023.acl-long.14 | A Method for Studying Semantic Construal in Grammatical Constructions with Interpretable Contextual Embedding Spaces | https://aclanthology.org/2023.acl-long.14/ | [
"Gabriella Chronis",
"Kyle Mahowald",
"Katrin Erk"
] | We study semantic construal in grammatical constructions using large language models. First, we project contextual word embeddings into three interpretable semantic spaces, each defined by a different set of psycholinguistic feature norms. We validate these interpretable spaces and then use them to automatically derive... | 2023.acl-long.14 | 10.18653/v1/2023.acl-long.14 | null | 2305.18598 | title_snapshot |
2023.acl-long.15 | Holographic CCG Parsing | https://aclanthology.org/2023.acl-long.15/ | [
"Ryosuke Yamaki",
"Tadahiro Taniguchi",
"Daichi Mochihashi"
] | We propose a method for formulating CCG as a recursive composition in a continuous vector space. Recent CCG supertagging and parsing models generally demonstrate high performance, yet rely on black-box neural architectures to implicitly model phrase structure dependencies. Instead, we leverage the method of holographic... | 2023.acl-long.15 | 10.18653/v1/2023.acl-long.15 | null | null | null |
2023.acl-long.16 | Prompts Can Play Lottery Tickets Well: Achieving Lifelong Information Extraction via Lottery Prompt Tuning | https://aclanthology.org/2023.acl-long.16/ | [
"Zujie Liang",
"Feng Wei",
"Yin Jie",
"Yuxi Qian",
"Zhenghong Hao",
"Bing Han"
] | Thanks to the recent success of Pre-trained Language Models (PLMs), it has become a promising research direction to develop a universal model (UIE) that can solve all typical information extraction tasks within one generative framework. Nonetheless, in real-world scenarios of UIE applications, new data of different IE ... | 2023.acl-long.16 | 10.18653/v1/2023.acl-long.16 | null | null | null |
2023.acl-long.17 | Retrieve-and-Sample: Document-level Event Argument Extraction via Hybrid Retrieval Augmentation | https://aclanthology.org/2023.acl-long.17/ | [
"Yubing Ren",
"Yanan Cao",
"Ping Guo",
"Fang Fang",
"Wei Ma",
"Zheng Lin"
] | Recent studies have shown the effectiveness of retrieval augmentation in many generative NLP tasks. These retrieval-augmented methods allow models to explicitly acquire prior external knowledge in a non-parametric manner and regard the retrieved reference instances as cues to augment text generation. These methods use ... | 2023.acl-long.17 | 10.18653/v1/2023.acl-long.17 | null | null | null |
2023.acl-long.18 | WeCheck: Strong Factual Consistency Checker via Weakly Supervised Learning | https://aclanthology.org/2023.acl-long.18/ | [
"Wenhao Wu",
"Wei Li",
"Xinyan Xiao",
"Jiachen Liu",
"Sujian Li",
"Yajuan Lyu"
] | A crucial issue of current text generation models is that they often uncontrollably generate text that is factually inconsistent with inputs. Due to lack of annotated data, existing factual consistency metrics usually train evaluation models on synthetic texts or directly transfer from other related tasks, such as ques... | 2023.acl-long.18 | 10.18653/v1/2023.acl-long.18 | null | 2212.10057 | title_snapshot |
2023.acl-long.19 | AMR-based Network for Aspect-based Sentiment Analysis | https://aclanthology.org/2023.acl-long.19/ | [
"Fukun Ma",
"Xuming Hu",
"Aiwei Liu",
"Yawen Yang",
"Shuang Li",
"Philip S. Yu",
"Lijie Wen"
] | Aspect-based sentiment analysis (ABSA) is a fine-grained sentiment classification task. Many recent works have used dependency trees to extract the relation between aspects and contexts and have achieved significant improvements. However, further improvement is limited due to the potential mismatch between the dependen... | 2023.acl-long.19 | 10.18653/v1/2023.acl-long.19 | null | null | null |
2023.acl-long.20 | Text Adversarial Purification as Defense against Adversarial Attacks | https://aclanthology.org/2023.acl-long.20/ | [
"Linyang Li",
"Demin Song",
"Xipeng Qiu"
] | Adversarial purification is a successful defense mechanism against adversarial attacks without requiring knowledge of the form of the incoming attack. Generally, adversarial purification aims to remove the adversarial perturbations therefore can make correct predictions based on the recovered clean samples. Despite the... | 2023.acl-long.20 | 10.18653/v1/2023.acl-long.20 | null | 2203.14207 | title_snapshot |
2023.acl-long.21 | SPEECH: Structured Prediction with Energy-Based Event-Centric Hyperspheres | https://aclanthology.org/2023.acl-long.21/ | [
"Shumin Deng",
"Shengyu Mao",
"Ningyu Zhang",
"Bryan Hooi"
] | Event-centric structured prediction involves predicting structured outputs of events. In most NLP cases, event structures are complex with manifold dependency, and it is challenging to effectively represent these complicated structured events. To address these issues, we propose Structured Prediction with Energy-based ... | 2023.acl-long.21 | 10.18653/v1/2023.acl-long.21 | null | 2305.13617 | title_snapshot |
2023.acl-long.22 | Rule By Example: Harnessing Logical Rules for Explainable Hate Speech Detection | https://aclanthology.org/2023.acl-long.22/ | [
"Christopher Clarke",
"Matthew Hall",
"Gaurav Mittal",
"Ye Yu",
"Sandra Sajeev",
"Jason Mars",
"Mei Chen"
] | Classic approaches to content moderation typically apply a rule-based heuristic approach to flag content. While rules are easily customizable and intuitive for humans to interpret, they are inherently fragile and lack the flexibility or robustness needed to moderate the vast amount of undesirable content found online t... | 2023.acl-long.22 | 10.18653/v1/2023.acl-long.22 | null | 2307.12935 | title_snapshot |
2023.acl-long.23 | What about “em”? How Commercial Machine Translation Fails to Handle (Neo-)Pronouns | https://aclanthology.org/2023.acl-long.23/ | [
"Anne Lauscher",
"Debora Nozza",
"Ehm Miltersen",
"Archie Crowley",
"Dirk Hovy"
] | As 3rd-person pronoun usage shifts to include novel forms, e.g., neopronouns, we need more research on identity-inclusive NLP. Exclusion is particularly harmful in one of the most popular NLP applications, machine translation (MT). Wrong pronoun translations can discriminate against marginalized groups, e.g., non-binar... | 2023.acl-long.23 | 10.18653/v1/2023.acl-long.23 | null | 2305.16051 | title_snapshot |
2023.acl-long.24 | What Is Overlap Knowledge in Event Argument Extraction? APE: A Cross-datasets Transfer Learning Model for EAE | https://aclanthology.org/2023.acl-long.24/ | [
"Kaihang Zhang",
"Kai Shuang",
"Xinyue Yang",
"Xuyang Yao",
"Jinyu Guo"
] | The EAE task extracts a structured event record from an event text. Most existing approaches train the EAE model on each dataset independently and ignore the overlap knowledge across datasets. However, insufficient event records in a single dataset often prevent the existing model from achieving better performance. In ... | 2023.acl-long.24 | 10.18653/v1/2023.acl-long.24 | null | null | null |
2023.acl-long.25 | Tailor: A Soft-Prompt-Based Approach to Attribute-Based Controlled Text Generation | https://aclanthology.org/2023.acl-long.25/ | [
"Kexin Yang",
"Dayiheng Liu",
"Wenqiang Lei",
"Baosong Yang",
"Mingfeng Xue",
"Boxing Chen",
"Jun Xie"
] | Attribute-based Controlled Text Generation (CTG) refers to generating sentences that satisfy desirable attributes (e.g., emotions and topics). Existing work usually utilize fine-tuning or resort to extra attribute classifiers, yet suffer from increases in storage and inference time. To address these concerns, we explor... | 2023.acl-long.25 | 10.18653/v1/2023.acl-long.25 | null | 2204.13362 | title_judge |
2023.acl-long.26 | Knowledge of cultural moral norms in large language models | https://aclanthology.org/2023.acl-long.26/ | [
"Aida Ramezani",
"Yang Xu"
] | Moral norms vary across cultures. A recent line of work suggests that English large language models contain human-like moral biases, but these studies typically do not examine moral variation in a diverse cultural setting. We investigate the extent to which monolingual English language models contain knowledge about mo... | 2023.acl-long.26 | 10.18653/v1/2023.acl-long.26 | null | 2306.01857 | title_snapshot |
2023.acl-long.27 | Songs Across Borders: Singable and Controllable Neural Lyric Translation | https://aclanthology.org/2023.acl-long.27/ | [
"Longshen Ou",
"Xichu Ma",
"Min-Yen Kan",
"Ye Wang"
] | The development of general-domain neural machine translation (NMT) methods has advanced significantly in recent years, but the lack of naturalness and musical constraints in the outputs makes them unable to produce singable lyric translations. This paper bridges the singability quality gap by formalizing lyric translat... | 2023.acl-long.27 | 10.18653/v1/2023.acl-long.27 | null | 2305.16816 | title_snapshot |
2023.acl-long.28 | Fantastic Expressions and Where to Find Them: Chinese Simile Generation with Multiple Constraints | https://aclanthology.org/2023.acl-long.28/ | [
"Kexin Yang",
"Dayiheng Liu",
"Wenqiang Lei",
"Baosong Yang",
"Xiangpeng Wei",
"Zhengyuan Liu",
"Jun Xie"
] | Similes occur in the creative context of describing a concept (i.e., tenor) by making a literally false yet figuratively meaningful comparison to another (i.e., vehicle). Previous efforts form simile generation as a context-free generation task, focusing on simile-style transfer or writing a simile from a given prefix.... | 2023.acl-long.28 | 10.18653/v1/2023.acl-long.28 | null | null | null |
2023.acl-long.29 | Revealing Single Frame Bias for Video-and-Language Learning | https://aclanthology.org/2023.acl-long.29/ | [
"Jie Lei",
"Tamara Berg",
"Mohit Bansal"
] | Training an effective video-and-language model intuitively requires multiple frames as model inputs. However, it is unclear whether using multiple frames is beneficial to downstream tasks, and if yes, whether the performance gain is worth the drastically-increased computation and memory costs resulting from using more ... | 2023.acl-long.29 | 10.18653/v1/2023.acl-long.29 | null | 2206.03428 | title_snapshot |
2023.acl-long.30 | Learning with Partial Annotations for Event Detection | https://aclanthology.org/2023.acl-long.30/ | [
"Jian Liu",
"Dianbo Sui",
"Kang Liu",
"Haoyan Liu",
"Zhe Zhao"
] | Event detection (ED) seeks to discover and classify event instances in plain texts. Previous methods for ED typically adopt supervised learning, requiring fully labeled and high-quality training data. However, in a real-world application, we may not obtain clean training data but only partially labeled one, which could... | 2023.acl-long.30 | 10.18653/v1/2023.acl-long.30 | null | null | null |
2023.acl-long.31 | World-to-Words: Grounded Open Vocabulary Acquisition through Fast Mapping in Vision-Language Models | https://aclanthology.org/2023.acl-long.31/ | [
"Ziqiao Ma",
"Jiayi Pan",
"Joyce Chai"
] | The ability to connect language units to their referents in the physical world, referred to as grounding, is crucial to learning and understanding grounded meanings of words. While humans demonstrate fast mapping in new word learning, it remains unclear whether modern vision-language models can truly represent language... | 2023.acl-long.31 | 10.18653/v1/2023.acl-long.31 | Outstanding Paper Award | 2306.08685 | title_snapshot |
2023.acl-long.32 | A Causal Framework to Quantify the Robustness of Mathematical Reasoning with Language Models | https://aclanthology.org/2023.acl-long.32/ | [
"Alessandro Stolfo",
"Zhijing Jin",
"Kumar Shridhar",
"Bernhard Schölkopf",
"Mrinmaya Sachan"
] | We have recently witnessed a number of impressive results on hard mathematical reasoning problems with language models. At the same time, the robustness of these models has also been called into question; recent works have shown that models can rely on shallow patterns in the problem description when generating a solut... | 2023.acl-long.32 | 10.18653/v1/2023.acl-long.32 | null | 2210.12023 | title_snapshot |
2023.acl-long.33 | Evaluating Open-Domain Dialogues in Latent Space with Next Sentence Prediction and Mutual Information | https://aclanthology.org/2023.acl-long.33/ | [
"Kun Zhao",
"Bohao Yang",
"Chenghua Lin",
"Wenge Rong",
"Aline Villavicencio",
"Xiaohui Cui"
] | The long-standing one-to-many issue of the open-domain dialogues poses significant challenges for automatic evaluation methods, i.e., there may be multiple suitable responses which differ in semantics for a given conversational context. To tackle this challenge, we propose a novel learning-based automatic evaluation me... | 2023.acl-long.33 | 10.18653/v1/2023.acl-long.33 | null | 2305.16967 | title_snapshot |
2023.acl-long.34 | Increasing Diversity While Maintaining Accuracy: Text Data Generation with Large Language Models and Human Interventions | https://aclanthology.org/2023.acl-long.34/ | [
"John Chung",
"Ece Kamar",
"Saleema Amershi"
] | Large language models (LLMs) can be used to generate text data for training and evaluating other models. However, creating high-quality datasets with LLMs can be challenging. In this work, we explore human-AI partnerships to facilitate high diversity and accuracy in LLM-based text data generation. We first examine two ... | 2023.acl-long.34 | 10.18653/v1/2023.acl-long.34 | null | 2306.04140 | title_snapshot |
2023.acl-long.35 | Pruning Pre-trained Language Models Without Fine-Tuning | https://aclanthology.org/2023.acl-long.35/ | [
"Ting Jiang",
"Deqing Wang",
"Fuzhen Zhuang",
"Ruobing Xie",
"Feng Xia"
] | To overcome the overparameterized problem in Pre-trained Language Models (PLMs), pruning is widely used as a simple and straightforward compression method by directly removing unimportant weights. Previous first-order methods successfully compress PLMs to extremely high sparsity with little performance drop. These meth... | 2023.acl-long.35 | 10.18653/v1/2023.acl-long.35 | null | 2210.06210 | title_snapshot |
2023.acl-long.36 | When Does Translation Require Context? A Data-driven, Multilingual Exploration | https://aclanthology.org/2023.acl-long.36/ | [
"Patrick Fernandes",
"Kayo Yin",
"Emmy Liu",
"André Martins",
"Graham Neubig"
] | Although proper handling of discourse significantly contributes to the quality of machine translation (MT), these improvements are not adequately measured in common translation quality metrics. Recent works in context-aware MT attempt to target a small set of discourse phenomena during evaluation, however not in a full... | 2023.acl-long.36 | 10.18653/v1/2023.acl-long.36 | Resource Award | 2109.07446 | title_snapshot |
2023.acl-long.37 | Causal Intervention and Counterfactual Reasoning for Multi-modal Fake News Detection | https://aclanthology.org/2023.acl-long.37/ | [
"Ziwei Chen",
"Linmei Hu",
"Weixin Li",
"Yingxia Shao",
"Liqiang Nie"
] | Due to the rapid upgrade of social platforms, most of today’s fake news is published and spread in a multi-modal form. Most existing multi-modal fake news detection methods neglect the fact that some label-specific features learned from the training set cannot generalize well to the testing set, thus inevitably sufferi... | 2023.acl-long.37 | 10.18653/v1/2023.acl-long.37 | null | null | null |
2023.acl-long.38 | LexSym: Compositionality as Lexical Symmetry | https://aclanthology.org/2023.acl-long.38/ | [
"Ekin Akyurek",
"Jacob Andreas"
] | In tasks like semantic parsing, instruction following, and question answering, standard deep networks fail to generalize compositionally from small datasets. Many existing approaches overcome this limitation with model architectures that enforce a compositional process of sentence interpretation. In this paper, we pres... | 2023.acl-long.38 | 10.18653/v1/2023.acl-long.38 | Area Chair Award (Semantics: Lexical) | null | null |
2023.acl-long.39 | Layer-wise Fusion with Modality Independence Modeling for Multi-modal Emotion Recognition | https://aclanthology.org/2023.acl-long.39/ | [
"Jun Sun",
"Shoukang Han",
"Yu-Ping Ruan",
"Xiaoning Zhang",
"Shu-Kai Zheng",
"Yulong Liu",
"Yuxin Huang",
"Taihao Li"
] | Multi-modal emotion recognition has gained increasing attention in recent years due to its widespread applications and the advances in multi-modal learning approaches. However, previous studies primarily focus on developing models that exploit the unification of multiple modalities. In this paper, we propose that maint... | 2023.acl-long.39 | 10.18653/v1/2023.acl-long.39 | null | null | null |
2023.acl-long.40 | CASN:Class-Aware Score Network for Textual Adversarial Detection | https://aclanthology.org/2023.acl-long.40/ | [
"Rong Bao",
"Rui Zheng",
"Liang Ding",
"Qi Zhang",
"Dacheng Tao"
] | Adversarial detection aims to detect adversarial samples that threaten the security of deep neural networks, which is an essential step toward building robust AI systems. Density-based estimation is widely considered as an effective technique by explicitly modeling the distribution of normal data and identifying advers... | 2023.acl-long.40 | 10.18653/v1/2023.acl-long.40 | null | null | null |
2023.acl-long.41 | Do Androids Laugh at Electric Sheep? Humor “Understanding” Benchmarks from The New Yorker Caption Contest | https://aclanthology.org/2023.acl-long.41/ | [
"Jack Hessel",
"Ana Marasovic",
"Jena D. Hwang",
"Lillian Lee",
"Jeff Da",
"Rowan Zellers",
"Robert Mankoff",
"Yejin Choi"
] | Large neural networks can now generate jokes, but do they really “understand” humor? We challenge AI models with three tasks derived from the New Yorker Cartoon Caption Contest: matching a joke to a cartoon, identifying a winning caption, and explaining why a winning caption is funny. These tasks encapsulate progressiv... | 2023.acl-long.41 | 10.18653/v1/2023.acl-long.41 | Best Paper Award | 2209.06293 | title_snapshot |
2023.acl-long.42 | Making More of Little Data: Improving Low-Resource Automatic Speech Recognition Using Data Augmentation | https://aclanthology.org/2023.acl-long.42/ | [
"Martijn Bartelds",
"Nay San",
"Bradley McDonnell",
"Dan Jurafsky",
"Martijn Wieling"
] | The performance of automatic speech recognition (ASR) systems has advanced substantially in recent years, particularly for languages for which a large amount of transcribed speech is available. Unfortunately, for low-resource languages, such as minority languages, regional languages or dialects, ASR performance general... | 2023.acl-long.42 | 10.18653/v1/2023.acl-long.42 | null | 2305.10951 | title_snapshot |
2023.acl-long.43 | CLCL: Non-compositional Expression Detection with Contrastive Learning and Curriculum Learning | https://aclanthology.org/2023.acl-long.43/ | [
"Jianing Zhou",
"Ziheng Zeng",
"Suma Bhat"
] | Non-compositional expressions present a substantial challenge for natural language processing (NLP) systems, necessitating more intricate processing compared to general language tasks, even with large pre-trained language models. Their non-compositional nature and limited availability of data resources further compound... | 2023.acl-long.43 | 10.18653/v1/2023.acl-long.43 | null | null | null |
2023.acl-long.44 | Multi-VALUE: A Framework for Cross-Dialectal English NLP | https://aclanthology.org/2023.acl-long.44/ | [
"Caleb Ziems",
"William Held",
"Jingfeng Yang",
"Jwala Dhamala",
"Rahul Gupta",
"Diyi Yang"
] | Dialect differences caused by regional, social, and economic factors cause performance discrepancies for many groups of language technology users. Inclusive and equitable language technology must critically be dialect invariant, meaning that performance remains constant over dialectal shifts. Current systems often fall... | 2023.acl-long.44 | 10.18653/v1/2023.acl-long.44 | null | 2212.08011 | title_snapshot |
2023.acl-long.45 | Self-Edit: Fault-Aware Code Editor for Code Generation | https://aclanthology.org/2023.acl-long.45/ | [
"Kechi Zhang",
"Zhuo Li",
"Jia Li",
"Ge Li",
"Zhi Jin"
] | Large language models (LLMs) have demonstrated an impressive ability to generate codes on competitive programming tasks. However, with limited sample numbers, LLMs still suffer from poor accuracy. Inspired by the process of human programming, we propose a generate-and-edit approach named Self-Edit that utilizes executi... | 2023.acl-long.45 | 10.18653/v1/2023.acl-long.45 | null | 2305.04087 | title_snapshot |
2023.acl-long.46 | ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning | https://aclanthology.org/2023.acl-long.46/ | [
"Shachar Don-Yehiya",
"Elad Venezian",
"Colin Raffel",
"Noam Slonim",
"Leshem Choshen"
] | Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now, massively multitask learning required simultaneous access to all datasets in the mixtur... | 2023.acl-long.46 | 10.18653/v1/2023.acl-long.46 | null | 2212.01378 | title_snapshot |
2023.acl-long.47 | Test-time Adaptation for Machine Translation Evaluation by Uncertainty Minimization | https://aclanthology.org/2023.acl-long.47/ | [
"Runzhe Zhan",
"Xuebo Liu",
"Derek F. Wong",
"Cuilian Zhang",
"Lidia S. Chao",
"Min Zhang"
] | The neural metrics recently received considerable attention from the research community in the automatic evaluation of machine translation. Unlike text-based metrics that have interpretable and consistent evaluation mechanisms for various data sources, the reliability of neural metrics in assessing out-of-distribution ... | 2023.acl-long.47 | 10.18653/v1/2023.acl-long.47 | null | null | null |
2023.acl-long.48 | Multi-CLS BERT: An Efficient Alternative to Traditional Ensembling | https://aclanthology.org/2023.acl-long.48/ | [
"Haw-Shiuan Chang",
"Ruei-Yao Sun",
"Kathryn Ricci",
"Andrew McCallum"
] | Ensembling BERT models often significantly improves accuracy, but at the cost of significantly more computation and memory footprint. In this work, we propose Multi-CLS BERT, a novel ensembling method for CLS-based prediction tasks that is almost as efficient as a single BERT model. Multi-CLS BERT uses multiple CLS tok... | 2023.acl-long.48 | 10.18653/v1/2023.acl-long.48 | null | 2210.05043 | title_snapshot |
2023.acl-long.49 | On-the-fly Cross-lingual Masking for Multilingual Pre-training | https://aclanthology.org/2023.acl-long.49/ | [
"Xi Ai",
"Bin Fang"
] | In multilingual pre-training with the objective of MLM (masked language modeling) on multiple monolingual corpora, multilingual models only learn cross-linguality implicitly from isomorphic spaces formed by overlapping different language spaces due to the lack of explicit cross-lingual forward pass. In this work, we pr... | 2023.acl-long.49 | 10.18653/v1/2023.acl-long.49 | null | null | null |
2023.acl-long.50 | How About Kind of Generating Hedges using End-to-End Neural Models? | https://aclanthology.org/2023.acl-long.50/ | [
"Alafate Abulimiti",
"Chloé Clavel",
"Justine Cassell"
] | Hedging is a strategy for softening the impact of a statement in conversation. In reducing the strength of an expression, it may help to avoid embarrassment (more technically, “face threat”) to one’s listener. For this reason, it is often found in contexts of instruction, such as tutoring. In this work, we develop a mo... | 2023.acl-long.50 | 10.18653/v1/2023.acl-long.50 | null | 2306.14696 | title_snapshot |
2023.acl-long.51 | DiffusionDB: A Large-scale Prompt Gallery Dataset for Text-to-Image Generative Models | https://aclanthology.org/2023.acl-long.51/ | [
"Zijie J. Wang",
"Evan Montoya",
"David Munechika",
"Haoyang Yang",
"Benjamin Hoover",
"Duen Horng Chau"
] | With recent advancements in diffusion models, users can generate high-quality images by writing text prompts in natural language. However, generating images with desired details requires proper prompts, and it is often unclear how a model reacts to different prompts or what the best prompts are. To help researchers tac... | 2023.acl-long.51 | 10.18653/v1/2023.acl-long.51 | null | 2210.14896 | title_snapshot |
2023.acl-long.52 | From Key Points to Key Point Hierarchy: Structured and Expressive Opinion Summarization | https://aclanthology.org/2023.acl-long.52/ | [
"Arie Cattan",
"Lilach Eden",
"Yoav Kantor",
"Roy Bar-Haim"
] | Key Point Analysis (KPA) has been recently proposed for deriving fine-grained insights from collections of textual comments. KPA extracts the main points in the data as a list of concise sentences or phrases, termed Key Points, and quantifies their prevalence. While key points are more expressive than word clouds and k... | 2023.acl-long.52 | 10.18653/v1/2023.acl-long.52 | null | 2306.03853 | title_snapshot |
2023.acl-long.53 | When to Use What: An In-Depth Comparative Empirical Analysis of OpenIE Systems for Downstream Applications | https://aclanthology.org/2023.acl-long.53/ | [
"Kevin Pei",
"Ishan Jindal",
"Kevin Chen-Chuan Chang",
"ChengXiang Zhai",
"Yunyao Li"
] | Open Information Extraction (OpenIE) has been used in the pipelines of various NLP tasks. Unfortunately, there is no clear consensus on which models to use in which tasks. Muddying things further is the lack of comparisons that take differing training sets into account. In this paper, we present an application-focused ... | 2023.acl-long.53 | 10.18653/v1/2023.acl-long.53 | null | 2211.08228 | title_snapshot |
2023.acl-long.54 | Subjective Crowd Disagreements for Subjective Data: Uncovering Meaningful CrowdOpinion with Population-level Learning | https://aclanthology.org/2023.acl-long.54/ | [
"Tharindu Cyril Weerasooriya",
"Sarah Luger",
"Saloni Poddar",
"Ashiqur KhudaBukhsh",
"Christopher Homan"
] | Human-annotated data plays a critical role in the fairness of AI systems, including those that deal with life-altering decisions or moderating human-created web/social media content. Conventionally, annotator disagreements are resolved before any learning takes place. However, researchers are increasingly identifying a... | 2023.acl-long.54 | 10.18653/v1/2023.acl-long.54 | null | 2307.10189 | title_snapshot |
2023.acl-long.55 | Post-Abstention: Towards Reliably Re-Attempting the Abstained Instances in QA | https://aclanthology.org/2023.acl-long.55/ | [
"Neeraj Varshney",
"Chitta Baral"
] | Despite remarkable progress made in natural language processing, even the state-of-the-art models often make incorrect predictions. Such predictions hamper the reliability of systems and limit their widespread adoption in real-world applications. ‘Selective prediction’ partly addresses the above concern by enabling mod... | 2023.acl-long.55 | 10.18653/v1/2023.acl-long.55 | null | 2305.01812 | title_snapshot |
2023.acl-long.56 | UniLG: A Unified Structure-aware Framework for Lyrics Generation | https://aclanthology.org/2023.acl-long.56/ | [
"Tao Qian",
"Fan Lou",
"Jiatong Shi",
"Yuning Wu",
"Shuai Guo",
"Xiang Yin",
"Qin Jin"
] | As a special task of natural language generation, conditional lyrics generation needs to consider the structure of generated lyrics and the relationship between lyrics and music. Due to various forms of conditions, a lyrics generation system is expected to generate lyrics conditioned on different signals, such as music... | 2023.acl-long.56 | 10.18653/v1/2023.acl-long.56 | null | null | null |
2023.acl-long.57 | FC-KBQA: A Fine-to-Coarse Composition Framework for Knowledge Base Question Answering | https://aclanthology.org/2023.acl-long.57/ | [
"Lingxi Zhang",
"Jing Zhang",
"Yanling Wang",
"Shulin Cao",
"Xinmei Huang",
"Cuiping Li",
"Hong Chen",
"Juanzi Li"
] | The generalization problem on KBQA has drawn considerable attention. Existing research suffers from the generalization issue brought by the entanglement in the coarse-grained modeling of the logical expression, or inexecutability issues due to the fine-grained modeling of disconnected classes and relations in real KBs.... | 2023.acl-long.57 | 10.18653/v1/2023.acl-long.57 | null | 2306.14722 | title_snapshot |
2023.acl-long.58 | Does GPT-3 Grasp Metaphors? Identifying Metaphor Mappings with Generative Language Models | https://aclanthology.org/2023.acl-long.58/ | [
"Lennart Wachowiak",
"Dagmar Gromann"
] | Conceptual metaphors present a powerful cognitive vehicle to transfer knowledge structures from a source to a target domain. Prior neural approaches focus on detecting whether natural language sequences are metaphoric or literal. We believe that to truly probe metaphoric knowledge in pre-trained language models, their ... | 2023.acl-long.58 | 10.18653/v1/2023.acl-long.58 | null | null | null |
2023.acl-long.59 | Being Right for Whose Right Reasons? | https://aclanthology.org/2023.acl-long.59/ | [
"Terne Sasha Thorn Jakobsen",
"Laura Cabello",
"Anders Søgaard"
] | Explainability methods are used to benchmark the extent to which model predictions align with human rationales i.e., are ‘right for the right reasons’. Previous work has failed to acknowledge, however, that what counts as a rationale is sometimes subjective. This paper presents what we think is a first of its kind, a c... | 2023.acl-long.59 | 10.18653/v1/2023.acl-long.59 | null | 2306.00639 | title_snapshot |
2023.acl-long.60 | ALERT: Adapt Language Models to Reasoning Tasks | https://aclanthology.org/2023.acl-long.60/ | [
"Ping Yu",
"Tianlu Wang",
"Olga Golovneva",
"Badr AlKhamissi",
"Siddharth Verma",
"Zhijing Jin",
"Gargi Ghosh",
"Mona Diab",
"Asli Celikyilmaz"
] | Recent advancements in large language models have enabled them to perform well on complex tasks that require step-by-step reasoning with few-shot learning. However, it is unclear whether these models are applying reasoning skills they have learnt during pre-training , or if they are simply memorizing their training cor... | 2023.acl-long.60 | 10.18653/v1/2023.acl-long.60 | null | null | null |
2023.acl-long.61 | Glot500: Scaling Multilingual Corpora and Language Models to 500 Languages | https://aclanthology.org/2023.acl-long.61/ | [
"Ayyoob Imani",
"Peiqin Lin",
"Amir Hossein Kargaran",
"Silvia Severini",
"Masoud Jalili Sabet",
"Nora Kassner",
"Chunlan Ma",
"Helmut Schmid",
"André Martins",
"François Yvon",
"Hinrich Schütze"
] | The NLP community has mainly focused on scaling Large Language Models (LLMs) vertically, i.e., making them better for about 100 languages. We instead scale LLMs horizontally: we create, through continued pretraining, Glot500-m, an LLM that covers 511 predominantly low-resource languages. An important part of this effor... | 2023.acl-long.61 | 10.18653/v1/2023.acl-long.61 | Area Chair Award (Multilingualism and Cross-Lingual NLP) | 2305.12182 | title_snapshot |
2023.acl-long.62 | Joint Constrained Learning with Boundary-adjusting for Emotion-Cause Pair Extraction | https://aclanthology.org/2023.acl-long.62/ | [
"Huawen Feng",
"Junlong Liu",
"Junhao Zheng",
"Haibin Chen",
"Xichen Shang",
"Qianli Ma"
] | Emotion-Cause Pair Extraction (ECPE) aims to identify the document’s emotion clauses and corresponding cause clauses. Like other relation extraction tasks, ECPE is closely associated with the relationship between sentences. Recent methods based on Graph Convolutional Networks focus on how to model the multiplex relatio... | 2023.acl-long.62 | 10.18653/v1/2023.acl-long.62 | null | null | null |
2023.acl-long.63 | Pretrained Bidirectional Distillation for Machine Translation | https://aclanthology.org/2023.acl-long.63/ | [
"Yimeng Zhuang",
"Mei Tu"
] | Knowledge transfer can boost neural machine translation (NMT), for example, by finetuning a pretrained masked language model (LM). However, it may suffer from the forgetting problem and the structural inconsistency between pretrained LMs and NMT models. Knowledge distillation (KD) may be a potential solution to allevia... | 2023.acl-long.63 | 10.18653/v1/2023.acl-long.63 | null | null | null |
2023.acl-long.64 | Pivotal Role of Language Modeling in Recommender Systems: Enriching Task-specific and Task-agnostic Representation Learning | https://aclanthology.org/2023.acl-long.64/ | [
"Kyuyong Shin",
"Hanock Kwak",
"Wonjae Kim",
"Jisu Jeong",
"Seungjae Jung",
"Kyungmin Kim",
"Jung-Woo Ha",
"Sang-Woo Lee"
] | Recent studies have proposed unified user modeling frameworks that leverage user behavior data from various applications. Many of them benefit from utilizing users’ behavior sequences as plain texts, representing rich information in any domain or system without losing generality. Hence, a question arises: Can language ... | 2023.acl-long.64 | 10.18653/v1/2023.acl-long.64 | null | 2212.03760 | title_snapshot |
2023.acl-long.65 | Improving Continual Relation Extraction by Distinguishing Analogous Semantics | https://aclanthology.org/2023.acl-long.65/ | [
"Wenzheng Zhao",
"Yuanning Cui",
"Wei Hu"
] | Continual relation extraction (RE) aims to learn constantly emerging relations while avoiding forgetting the learned relations. Existing works store a small number of typical samples to re-train the model for alleviating forgetting. However, repeatedly replaying these samples may cause the overfitting problem. We condu... | 2023.acl-long.65 | 10.18653/v1/2023.acl-long.65 | null | 2305.06620 | title_snapshot |
2023.acl-long.66 | Improving Pretraining Techniques for Code-Switched NLP | https://aclanthology.org/2023.acl-long.66/ | [
"Richeek Das",
"Sahasra Ranjan",
"Shreya Pathak",
"Preethi Jyothi"
] | Pretrained models are a mainstay in modern NLP applications. Pretraining requires access to large volumes of unlabeled text. While monolingual text is readily available for many of the world’s languages, access to large quantities of code-switched text (i.e., text with tokens of multiple languages interspersed within a... | 2023.acl-long.66 | 10.18653/v1/2023.acl-long.66 | Outstanding Paper Award | null | null |
2023.acl-long.67 | A Theory of Unsupervised Speech Recognition | https://aclanthology.org/2023.acl-long.67/ | [
"Liming Wang",
"Mark Hasegawa-Johnson",
"Chang Yoo"
] | Unsupervised speech recognition ({pasted macro ‘ASRU’}/) is the problem of learning automatic speech recognition (ASR) systems from unpaired speech-only and text-only corpora. While various algorithms exist to solve this problem, a theoretical framework is missing to study their properties and address such issues as se... | 2023.acl-long.67 | 10.18653/v1/2023.acl-long.67 | null | 2306.07926 | title_snapshot |
2023.acl-long.68 | ThinkSum: Probabilistic reasoning over sets using large language models | https://aclanthology.org/2023.acl-long.68/ | [
"Batu Ozturkler",
"Nikolay Malkin",
"Zhen Wang",
"Nebojsa Jojic"
] | Large language models (LLMs) have a substantial capacity for high-level analogical reasoning: reproducing patterns in linear text that occur in their training data (zero-shot evaluation) or in the provided context (few-shot in-context learning). However, recent studies show that even the more advanced LLMs fail in scen... | 2023.acl-long.68 | 10.18653/v1/2023.acl-long.68 | null | 2210.01293 | title_snapshot |
2023.acl-long.69 | NLG Evaluation Metrics Beyond Correlation Analysis: An Empirical Metric Preference Checklist | https://aclanthology.org/2023.acl-long.69/ | [
"Iftitahu Nimah",
"Meng Fang",
"Vlado Menkovski",
"Mykola Pechenizkiy"
] | In this study, we analyze automatic evaluation metrics for Natural Language Generation (NLG), specifically task-agnostic metrics and human-aligned metrics. Task-agnostic metrics, such as Perplexity, BLEU, BERTScore, are cost-effective and highly adaptable to diverse NLG tasks, yet they have a weak correlation with huma... | 2023.acl-long.69 | 10.18653/v1/2023.acl-long.69 | null | 2305.08566 | title_snapshot |
2023.acl-long.70 | DialoGPS: Dialogue Path Sampling in Continuous Semantic Space for Data Augmentation in Multi-Turn Conversations | https://aclanthology.org/2023.acl-long.70/ | [
"Ang Lv",
"Jinpeng Li",
"Yuhan Chen",
"Gao Xing",
"Ji Zhang",
"Rui Yan"
] | In open-domain dialogue generation tasks, contexts and responses in most datasets are one-to-one mapped, violating an important many-to-many characteristic: a context leads to various responses, and a response answers multiple contexts. Without such patterns, models poorly generalize and prefer responding safely. Many ... | 2023.acl-long.70 | 10.18653/v1/2023.acl-long.70 | null | 2306.16770 | title_snapshot |
2023.acl-long.71 | TECHS: Temporal Logical Graph Networks for Explainable Extrapolation Reasoning | https://aclanthology.org/2023.acl-long.71/ | [
"Qika Lin",
"Jun Liu",
"Rui Mao",
"Fangzhi Xu",
"Erik Cambria"
] | Extrapolation reasoning on temporal knowledge graphs (TKGs) aims to forecast future facts based on past counterparts. There are two main challenges: (1) incorporating the complex information, including structural dependencies, temporal dynamics, and hidden logical rules; (2) implementing differentiable logical rule lea... | 2023.acl-long.71 | 10.18653/v1/2023.acl-long.71 | null | null | null |
2023.acl-long.72 | Consistency Regularization Training for Compositional Generalization | https://aclanthology.org/2023.acl-long.72/ | [
"Yongjing Yin",
"Jiali Zeng",
"Yafu Li",
"Fandong Meng",
"Jie Zhou",
"Yue Zhang"
] | Existing neural models have difficulty generalizing to unseen combinations of seen components. To achieve compositional generalization, models are required to consistently interpret (sub)expressions across contexts. Without modifying model architectures, we improve the capability of Transformer on compositional general... | 2023.acl-long.72 | 10.18653/v1/2023.acl-long.72 | null | null | null |
2023.acl-long.73 | NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generation | https://aclanthology.org/2023.acl-long.73/ | [
"Shengming Yin",
"Chenfei Wu",
"Huan Yang",
"Jianfeng Wang",
"Xiaodong Wang",
"Minheng Ni",
"Zhengyuan Yang",
"Linjie Li",
"Shuguang Liu",
"Fan Yang",
"Jianlong Fu",
"Ming Gong",
"Lijuan Wang",
"Zicheng Liu",
"Houqiang Li",
"Nan Duan"
] | In this paper, we propose NUWA-XL, a novel Diffusion over Diffusion architecture for eXtremely Long video generation. Most current work generates long videos segment by segment sequentially, which normally leads to the gap between training on short videos and inferring long videos, and the sequential generation is inef... | 2023.acl-long.73 | 10.18653/v1/2023.acl-long.73 | null | 2303.12346 | title_snapshot |
2023.acl-long.74 | Synthetic Text Generation with Differential Privacy: A Simple and Practical Recipe | https://aclanthology.org/2023.acl-long.74/ | [
"Xiang Yue",
"Huseyin Inan",
"Xuechen Li",
"Girish Kumar",
"Julia McAnallen",
"Hoda Shajari",
"Huan Sun",
"David Levitan",
"Robert Sim"
] | Privacy concerns have attracted increasing attention in data-driven products due to the tendency of machine learning models to memorize sensitive training data. Generating synthetic versions of such data with a formal privacy guarantee, such as differential privacy (DP), provides a promising path to mitigating these pr... | 2023.acl-long.74 | 10.18653/v1/2023.acl-long.74 | null | 2210.14348 | title_snapshot |
2023.acl-long.75 | A Close Look into the Calibration of Pre-trained Language Models | https://aclanthology.org/2023.acl-long.75/ | [
"Yangyi Chen",
"Lifan Yuan",
"Ganqu Cui",
"Zhiyuan Liu",
"Heng Ji"
] | Pre-trained language models (PLMs) may fail in giving reliable estimates of their predictive uncertainty. We take a close look into this problem, aiming to answer two questions: (1) Do PLMs learn to become calibrated in the training process? (2) How effective are existing calibration methods? For the first question, we... | 2023.acl-long.75 | 10.18653/v1/2023.acl-long.75 | null | 2211.00151 | title_snapshot |
2023.acl-long.76 | DIONYSUS: A Pre-trained Model for Low-Resource Dialogue Summarization | https://aclanthology.org/2023.acl-long.76/ | [
"Yu Li",
"Baolin Peng",
"Pengcheng He",
"Michel Galley",
"Zhou Yu",
"Jianfeng Gao"
] | Dialogue summarization has recently garnered significant attention due to its wide range of applications. However, existing methods for summarizing dialogues have limitations because they do not take into account the inherent structure of dialogue and rely heavily on labeled data, which can lead to poor performance in ... | 2023.acl-long.76 | 10.18653/v1/2023.acl-long.76 | null | 2212.10018 | title_snapshot |
2023.acl-long.77 | MS-DETR: Natural Language Video Localization with Sampling Moment-Moment Interaction | https://aclanthology.org/2023.acl-long.77/ | [
"Wang Jing",
"Aixin Sun",
"Hao Zhang",
"Xiaoli Li"
] | Given a text query, the task of Natural Language Video Localization (NLVL) is to localize a temporal moment in an untrimmed video that semantically matches the query. In this paper, we adopt a proposal-based solution that generates proposals (i.e. candidate moments) and then select the best matching proposal. On top of... | 2023.acl-long.77 | 10.18653/v1/2023.acl-long.77 | null | 2305.18969 | title_snapshot |
2023.acl-long.78 | Diverse Demonstrations Improve In-context Compositional Generalization | https://aclanthology.org/2023.acl-long.78/ | [
"Itay Levy",
"Ben Bogin",
"Jonathan Berant"
] | In-context learning has shown great success in i.i.d semantic parsing splits, where the training and test sets are drawn from the same distribution. In this setup, models are typically prompted with demonstrations that are similar to the input utterance. However, in the setup of compositional generalization, where mode... | 2023.acl-long.78 | 10.18653/v1/2023.acl-long.78 | null | 2212.06800 | title_snapshot |
2023.acl-long.79 | Self-Adaptive In-Context Learning: An Information Compression Perspective for In-Context Example Selection and Ordering | https://aclanthology.org/2023.acl-long.79/ | [
"Zhiyong Wu",
"Yaoxiang Wang",
"Jiacheng Ye",
"Lingpeng Kong"
] | Despite the surprising few-shot performance of in-context learning (ICL), it is still a common practice to randomly sample examples to serve as context. This paper advocates a new principle for ICL: self-adaptive in-context learning. The self-adaption mechanism is introduced to help each sample find an in-context examp... | 2023.acl-long.79 | 10.18653/v1/2023.acl-long.79 | null | 2212.10375 | title_snapshot |
2023.acl-long.80 | On the Efficacy of Sampling Adapters | https://aclanthology.org/2023.acl-long.80/ | [
"Clara Meister",
"Tiago Pimentel",
"Luca Malagutti",
"Ethan Wilcox",
"Ryan Cotterell"
] | Sampling-based decoding strategies are widely employed for generating text from probabilistic models, yet standard ancestral sampling often results in text that is degenerate or incoherent. To alleviate this issue, various modifications to a model’s sampling distribution, such as nucleus or top-k sampling, have been in... | 2023.acl-long.80 | 10.18653/v1/2023.acl-long.80 | null | 2307.03749 | title_snapshot |
2023.acl-long.81 | Cross-Domain Data Augmentation with Domain-Adaptive Language Modeling for Aspect-Based Sentiment Analysis | https://aclanthology.org/2023.acl-long.81/ | [
"Jianfei Yu",
"Qiankun Zhao",
"Rui Xia"
] | Cross-domain Aspect-Based Sentiment Analysis (ABSA) aims to leverage the useful knowledge from a source domain to identify aspect-sentiment pairs in sentences from a target domain. To tackle the task, several recent works explore a new unsupervised domain adaptation framework, i.e., Cross-Domain Data Augmentation (CDDA... | 2023.acl-long.81 | 10.18653/v1/2023.acl-long.81 | null | null | null |
2023.acl-long.82 | Compositional Data Augmentation for Abstractive Conversation Summarization | https://aclanthology.org/2023.acl-long.82/ | [
"Siru Ouyang",
"Jiaao Chen",
"Jiawei Han",
"Diyi Yang"
] | Recent abstractive conversation summarization systems generally rely on large-scale datasets with annotated summaries. However, collecting and annotating these conversations can be a time-consuming and labor-intensive task. To address this issue, in this work, we present a sub-structure level compositional data augment... | 2023.acl-long.82 | 10.18653/v1/2023.acl-long.82 | null | null | null |
2023.acl-long.83 | PMAES: Prompt-mapping Contrastive Learning for Cross-prompt Automated Essay Scoring | https://aclanthology.org/2023.acl-long.83/ | [
"Yuan Chen",
"Xia Li"
] | Current cross-prompt automated essay scoring (AES) is a challenging task due to the large discrepancies between different prompts, such as different genres and expressions. The main goal of current cross-prompt AES systems is to learn enough shared features between the source and target prompts to grade well on the tar... | 2023.acl-long.83 | 10.18653/v1/2023.acl-long.83 | null | null | null |
2023.acl-long.84 | Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Models | https://aclanthology.org/2023.acl-long.84/ | [
"Myra Cheng",
"Esin Durmus",
"Dan Jurafsky"
] | To recognize and mitigate harms from large language models (LLMs), we need to understand the prevalence and nuances of stereotypes in LLM outputs. Toward this end, we present Marked Personas, a prompt-based method to measure stereotypes in LLMs for intersectional demographic groups without any lexicon or data labeling.... | 2023.acl-long.84 | 10.18653/v1/2023.acl-long.84 | Social Impact Award | 2305.18189 | title_snapshot |
2023.acl-long.85 | On Prefix-tuning for Lightweight Out-of-distribution Detection | https://aclanthology.org/2023.acl-long.85/ | [
"Yawen Ouyang",
"Yongchang Cao",
"Yuan Gao",
"Zhen Wu",
"Jianbing Zhang",
"Xinyu Dai"
] | Out-of-distribution (OOD) detection, a fundamental task vexing real-world applications, has attracted growing attention in the NLP community. Recently fine-tuning based methods have made promising progress. However, it could be costly to store fine-tuned models for each scenario. In this paper, we depart from the class... | 2023.acl-long.85 | 10.18653/v1/2023.acl-long.85 | null | null | null |
2023.acl-long.86 | GEC-DePenD: Non-Autoregressive Grammatical Error Correction with Decoupled Permutation and Decoding | https://aclanthology.org/2023.acl-long.86/ | [
"Konstantin Yakovlev",
"Alexander Podolskiy",
"Andrey Bout",
"Sergey Nikolenko",
"Irina Piontkovskaya"
] | Grammatical error correction (GEC) is an important NLP task that is currently usually solved with autoregressive sequence-to-sequence models. However, approaches of this class are inherently slow due to one-by-one token generation, so non-autoregressive alternatives are needed. In this work, we propose a novel non-auto... | 2023.acl-long.86 | 10.18653/v1/2023.acl-long.86 | null | 2311.08191 | title_snapshot |
2023.acl-long.87 | Measuring Progress in Fine-grained Vision-and-Language Understanding | https://aclanthology.org/2023.acl-long.87/ | [
"Emanuele Bugliarello",
"Laurent Sartran",
"Aishwarya Agrawal",
"Lisa Anne Hendricks",
"Aida Nematzadeh"
] | While pretraining on large-scale image–text data from the Web has facilitated rapid progress on many vision-and-language (V&L) tasks, recent work has demonstrated that pretrained models lack “fine-grained” understanding, such as the ability to recognise relationships, verbs, and numbers in images. This has resulted in ... | 2023.acl-long.87 | 10.18653/v1/2023.acl-long.87 | null | 2305.07558 | title_snapshot |
2023.acl-long.88 | Vision Meets Definitions: Unsupervised Visual Word Sense Disambiguation Incorporating Gloss Information | https://aclanthology.org/2023.acl-long.88/ | [
"Sunjae Kwon",
"Rishabh Garodia",
"Minhwa Lee",
"Zhichao Yang",
"Hong Yu"
] | Visual Word Sense Disambiguation (VWSD) is a task to find the image that most accurately depicts the correct sense of the target word for the given context. Previously, image-text matching models often suffered from recognizing polysemous words. This paper introduces an unsupervised VWSD approach that uses gloss inform... | 2023.acl-long.88 | 10.18653/v1/2023.acl-long.88 | null | 2305.01788 | title_snapshot |
2023.acl-long.89 | Chain-of-Skills: A Configurable Model for Open-Domain Question Answering | https://aclanthology.org/2023.acl-long.89/ | [
"Kaixin Ma",
"Hao Cheng",
"Yu Zhang",
"Xiaodong Liu",
"Eric Nyberg",
"Jianfeng Gao"
] | The retrieval model is an indispensable component for real-world knowledge-intensive tasks, e.g., open-domain question answering (ODQA). As separate retrieval skills are annotated for different datasets, recent work focuses on customized methods, limiting the model transfer- ability and scalability. In this work, we pr... | 2023.acl-long.89 | 10.18653/v1/2023.acl-long.89 | null | 2305.03130 | title_snapshot |
2023.acl-long.90 | Elaboration-Generating Commonsense Question Answering at Scale | https://aclanthology.org/2023.acl-long.90/ | [
"Wenya Wang",
"Vivek Srikumar",
"Hannaneh Hajishirzi",
"Noah A. Smith"
] | In question answering requiring common sense, language models (e.g., GPT-3) have been used to generate text expressing background knowledge that helps improve performance. Yet the cost of working with such models is very high; in this work, we finetune smaller language models to generate useful intermediate context, re... | 2023.acl-long.90 | 10.18653/v1/2023.acl-long.90 | null | 2209.01232 | title_snapshot |
2023.acl-long.91 | Neural Unsupervised Reconstruction of Protolanguage Word Forms | https://aclanthology.org/2023.acl-long.91/ | [
"Andre He",
"Nicholas Tomlin",
"Dan Klein"
] | We present a state-of-the-art neural approach to the unsupervised reconstruction of ancient word forms. Previous work in this domain used expectation-maximization to predict simple phonological changes between ancient word forms and their cognates in modern languages. We extend this work with neural models that can cap... | 2023.acl-long.91 | 10.18653/v1/2023.acl-long.91 | null | 2211.08684 | title_snapshot |
2023.acl-long.92 | DaMSTF: Domain Adversarial Learning Enhanced Meta Self-Training for Domain Adaptation | https://aclanthology.org/2023.acl-long.92/ | [
"Menglong Lu",
"Zhen Huang",
"Yunxiang Zhao",
"Zhiliang Tian",
"Yang Liu",
"Dongsheng Li"
] | Self-training emerges as an important research line on domain adaptation. By taking the model’s prediction as the pseudo labels of the unlabeled data, self-training bootstraps the model with pseudo instances in the target domain. However, the prediction errors of pseudo labels (label noise) challenge the performance of... | 2023.acl-long.92 | 10.18653/v1/2023.acl-long.92 | null | 2308.02753 | title_snapshot |
2023.acl-long.93 | On Evaluating Multilingual Compositional Generalization with Translated Datasets | https://aclanthology.org/2023.acl-long.93/ | [
"Zi Wang",
"Daniel Hershcovich"
] | Compositional generalization allows efficient learning and human-like inductive biases. Since most research investigating compositional generalization in NLP is done on English, important questions remain underexplored. Do the necessary compositional generalization abilities differ across languages? Can models composit... | 2023.acl-long.93 | 10.18653/v1/2023.acl-long.93 | null | 2306.11420 | title_snapshot |
2023.acl-long.94 | FAA: Fine-grained Attention Alignment for Cascade Document Ranking | https://aclanthology.org/2023.acl-long.94/ | [
"Zhen Li",
"Chongyang Tao",
"Jiazhan Feng",
"Tao Shen",
"Dongyan Zhao",
"Xiubo Geng",
"Daxin Jiang"
] | Document ranking aims at sorting a collection of documents with their relevance to a query. Contemporary methods explore more efficient transformers or divide long documents into passages to handle the long input. However, intensive query-irrelevant content may lead to harmful distraction and high query latency. Some r... | 2023.acl-long.94 | 10.18653/v1/2023.acl-long.94 | null | null | null |
2023.acl-long.95 | Fine-tuning Happens in Tiny Subspaces: Exploring Intrinsic Task-specific Subspaces of Pre-trained Language Models | https://aclanthology.org/2023.acl-long.95/ | [
"Zhong Zhang",
"Bang Liu",
"Junming Shao"
] | Pre-trained language models (PLMs) are known to be overly parameterized and have significant redundancy, indicating a small degree of freedom of the PLMs. Motivated by the observation, in this paper, we study the problem of re-parameterizing and fine-tuning PLMs from a new perspective: Discovery of intrinsic task-speci... | 2023.acl-long.95 | 10.18653/v1/2023.acl-long.95 | null | 2305.17446 | title_snapshot |
2023.acl-long.96 | Facilitating Multi-turn Emotional Support Conversation with Positive Emotion Elicitation: A Reinforcement Learning Approach | https://aclanthology.org/2023.acl-long.96/ | [
"Jinfeng Zhou",
"Zhuang Chen",
"Bo Wang",
"Minlie Huang"
] | Emotional support conversation (ESC) aims to provide emotional support (ES) to improve one’s mental state. Existing works stay at fitting grounded responses and responding strategies (e.g., question), which ignore the effect on ES and lack explicit goals to guide emotional positive transition. To this end, we introduce... | 2023.acl-long.96 | 10.18653/v1/2023.acl-long.96 | null | 2307.07994 | title_snapshot |
2023.acl-long.97 | Query Enhanced Knowledge-Intensive Conversation via Unsupervised Joint Modeling | https://aclanthology.org/2023.acl-long.97/ | [
"Mingzhu Cai",
"Siqi Bao",
"Xin Tian",
"Huang He",
"Fan Wang",
"Hua Wu"
] | In this paper, we propose an unsupervised query enhanced approach for knowledge-intensive conversations, namely QKConv. There are three modules in QKConv: a query generator, an off-the-shelf knowledge selector, and a response generator. QKConv is optimized through joint training, which produces the response by explorin... | 2023.acl-long.97 | 10.18653/v1/2023.acl-long.97 | null | 2212.09588 | title_snapshot |
2023.acl-long.98 | Why Aren’t We NER Yet? Artifacts of ASR Errors in Named Entity Recognition in Spontaneous Speech Transcripts | https://aclanthology.org/2023.acl-long.98/ | [
"Piotr Szymański",
"Lukasz Augustyniak",
"Mikolaj Morzy",
"Adrian Szymczak",
"Krzysztof Surdyk",
"Piotr Żelasko"
] | Transcripts of spontaneous human speech present a significant obstacle for traditional NER models. The lack of grammatical structure of spoken utterances and word errors introduced by the ASR make downstream NLP tasks challenging. In this paper, we examine in detail the complex relationship between ASR and NER errors w... | 2023.acl-long.98 | 10.18653/v1/2023.acl-long.98 | null | null | null |
2023.acl-long.99 | Precise Zero-Shot Dense Retrieval without Relevance Labels | https://aclanthology.org/2023.acl-long.99/ | [
"Luyu Gao",
"Xueguang Ma",
"Jimmy Lin",
"Jamie Callan"
] | While dense retrieval has been shown to be effective and efficient across tasks and languages, it remains difficult to create effective fully zero-shot dense retrieval systems when no relevance labels are available. In this paper, we recognize the difficulty of zero-shot learning and encoding relevance. Instead, we pro... | 2023.acl-long.99 | 10.18653/v1/2023.acl-long.99 | null | 2212.10496 | title_snapshot |
2023.acl-long.100 | White-Box Multi-Objective Adversarial Attack on Dialogue Generation | https://aclanthology.org/2023.acl-long.100/ | [
"Yufei Li",
"Zexin Li",
"Yingfan Gao",
"Cong Liu"
] | Pre-trained transformers are popular in state-of-the-art dialogue generation (DG) systems. Such language models are, however, vulnerable to various adversarial samples as studied in traditional tasks such as text classification, which inspires our curiosity about their robustness in DG systems. One main challenge of at... | 2023.acl-long.100 | 10.18653/v1/2023.acl-long.100 | null | 2305.03655 | title_snapshot |
End of preview. Expand in Data Studio
README.md exists but content is empty.
- Downloads last month
- -