new

Get trending papers in your email inbox once a day!

Get trending papers in your email inbox!

Daily Papers

byAK and the research community

July

Submitted by

taesiri

Kimi K3: Open Frontier Intelligence

moonshotai

Submitted by

yh-wang

Orca: The World is in Your Mind

·
57 authors

Submitted by

taesiri

ABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPU

acvlab

Alibaba AMAP CV Lab

Submitted by

kongquyu

Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents

AlibabaTongyiLab

7

Submitted by

yuntian-deng

Program-as-Weights: A Programming Paradigm for Fuzzy Functions

UWaterloo

University of Waterloo

Submitted by

bing-yan

AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis

newyorkuniversity

New York University

Submitted by

astarkkk

Metis: Memory Foundation Model

MemTensor

Submitted by

taesiri

Harness Handbook: Making Evolving Agent Harnesses Readable,Navigable, and Editable

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

xtma

LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget

mindlab-research

Submitted by

huangsiteng

RynnBrain 1.1: Towards More Capable and Generalizable Embodied Foundation Model

Alibaba-DAMO-Academy

3

Submitted by

Sterzhang

Progress Reward Modeling for Robotic Learning: A Comprehensive Survey

northwestern-university

Northwestern University

Submitted by

iseesaw

Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering

FrontisAI

Submitted by

taesiri

PhiZero: A World Model Built Around Physical Language

CASIA

Chinese Academic of Science Institute of Automation

Submitted by

lmwang

VideoChat3: Fully Open Video MLLM for Efficient and Generalist Video Understanding

MCG-NJU

Multimedia Computing Group-Nanjing University

Submitted by

AnitaLeung

The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning

·
12 authors

Submitted by

Lanxingxuan

TimeLens2: Generalist Video Temporal Grounding with Multimodal LLMs

MCG-NJU

Multimedia Computing Group-Nanjing University

Submitted by

Jarvis1111

HiFi-UMI: Learning Deployable Manipulation Policies from High-Fidelity UMI Data Alone

simple-world-lab

Simple World Lab

2

Submitted by

lz1001

AREX: Towards a Recursively Self-Improving Agent for Deep Research

BAAI

Beijing Academy of Artificial Intelligence

Submitted by

bond005

RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM

NSU

Novosibirsk State University

Submitted by

c7w

Weak-to-Strong Generalization via Direct On-Policy Distillation

BytedTsinghua-SIA

BytedTsinghua-SIA

Submitted by

jt-zhang

Vidu S1: A Real-Time Interactive Video Generation Model

Tsinghua University

Submitted by

lhpku20010120

DataFlow-Harness: A Grounded Code-Agent Platform for Constructing Editable LLM Data Pipelines

PekingUniversity

Peking University

2

Submitted by

Yif29

RESOURCE2SKILL: Distilling Executable Agent Skills from Human-Created Multimodal Resources

MicrosoftResearch

Microsoft Research

Submitted by

Chufeng

Boogu-Image-0.1: Boosting Open Agentic Multimodal Generation via Understanding under a Minimal Budget

Boogu

Submitted by

dkliang

TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM

H-EmbodVis

Submitted by

LYL1015

JarvisHub: An Open Harness for Canvas-Native Multimodal Creative Agents

·
26 authors

Submitted by

YerbaPage

Dockerless: Environment-Free Program Verifier for Coding Agents

ByteDance

3

Submitted by

yxl66666

DOPD: Dual On-policy Distillation

·
16 authors

2

Submitted by

fishmingyu

CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents

sysevol-ai

SysEvol AI Research

Submitted by

ubowang

Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models

TIGER-Lab

Submitted by

Jinyang23

SEED: Self-Evolving On-Policy Distillation for Agentic Reinforcement Learning

·
11 authors

Submitted by

mingchenlin2025

ABot-N1: Toward a General Visual Language Navigation Foundation Model

acvlab

Alibaba AMAP CV Lab

Submitted by

txy

Ring-Zero: Scaling Zero RL to a Trillion Parameters for Emergent Reasoning

RUC-AIBOX

Submitted by

MindscapeRAG

A New Role for Relevance: Guiding Corpus Interaction in Agentic Search

tencent

Submitted by

taesiri

RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation

·
6 authors

Submitted by

Ornamentt

DeepSearch-World: Self-Distillation for Deep Search Agents in a Verifiable Environment

Submitted by

LiJiaxing

DistillAlign: Coordinating Mode Covering and Mode Seeking in Autoregressive Video Distillation

·
10 authors

Submitted by

JiayuJeff

EvolvingWorld: An Open-Schema Framework for Co-Evolving Role-Play Agents and World Model in Interactive Literary World

tencent

Submitted by

taesiri

AlayaWorld: Long-Horizon and Playable Video World Generation

·
17 authors

Submitted by

taesiri

Video Generation Models are General-Purpose Vision Learners

GoogleDeepMind

Submitted by

Uanu

Accurate, Interdisciplinary and Transparent Structure-property Understanding with Deep Native Structural Reasoning

Shanghai-AI-Laboratory

Shanghai AI Laboratory

Submitted by

CongWei1230

Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation

TIGER-Lab

Submitted by

huangrh9

Read It Back: Pretrained MLLMs Are Zero-Shot Reward Models for Text-to-Image Generation

ByteDance-Seed

3

Submitted by

CheeryLJH

ABot-AgentOS: A General Robotic Agent OS with Lifelong Multi-modal Memory

acvlab

Alibaba AMAP CV Lab

Submitted by

jwhe

CoRT: Counterfactual Replay for Token-Level Rubric-Guided Policy Optimization

ByteDance

3

Submitted by

kailinjiang

From Proprietary to Open-Source: Bridging the Distribution Gap via Multi-Agent Protocol Distillation in Agentic Search

·
10 authors

2

Submitted by

rayleizhu

Hierarchical Sparse Attention Done Right: Toward Infinite Context Modeling

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

wangzx1994

Generative World Renderer at the Speed of Play

AlayaLab

Submitted by

osanseviero

Gemma 4 Technical Report

google

5

Submitted by

taesiri

SWE-Pruner Pro: The Coder LLM Already Knows What to Prune

ByteDance

Submitted by

YunxinLi

KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and Skill

HIT-TMG

Submitted by

Cuttle-fish-my

Rethinking Classifier-Free Guidance in On-Policy Diffusion Distillation

QwenBusinessUnit

Qwen Business Unit

Submitted by

nielsr

MonkeyOCRv2: A Visual-Text Foundation Model for Document AI

Submitted by

tttonyalpha

Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models

·
13 authors

Submitted by

taesiri

RynnWorld-Teleop: An Action-Conditioned World Model for Digital Teleoperation

·
9 authors

Submitted by

kuvvi

HumanCLAW: Can Vision-Language Models Act Through a Body?

metaresearch

Submitted by

Xinjie-Q

Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing

microsoft

2

Submitted by

taesiri

Loop the Loopies!

IQuestLab

Submitted by

zli12321

Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

chengtan9907

OmniOpt: Taxonomy, Geometry, and Benchmarking of Modern Optimizers

opendatalab

Submitted by

seashell11

BlockPilot: Instance-Adaptive Policy Learning for Diffusion-based Speculative Decoding

·
6 authors

Submitted by

crazyofapple

SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD

·
65 authors

Submitted by

Met4physics

Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers

RTP-LLM

Submitted by

YeWen27

Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories

XiaomiRobotics

Xiaomi Robotics

Submitted by

BroAlanTaps

UI-MOPD: Multi-Platform On-Policy Distillation for Continual GUI Agent Learning

·
11 authors

Submitted by

Lin-Chen

VideoCoCo: Code-as-CoT for Physically-Consistent Video Generation via an Agentic Dual-Engine System

·
28 authors

Submitted by

taesiri

SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration

antgroup

Submitted by

amitsaha

SynthDocBench: Controlled Benchmark for Long-Context Visual Document Understanding

ServiceNow-AI

Submitted by

kpzhang996

AgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM Agents

AlayaLab

Submitted by

jwchen2001

DecoEvo: Score-Decoupled Co-Evolution of Solver and Rubric-Generator Skills in Text Space

QwenBusinessUnit

Qwen Business Unit

3

Submitted by

Woxxxe

Open-AoE: An Open Egocentric Manipulation Dataset and Toolchain for Embodied Learning

inclusionAI

Submitted by

YeolJoo

ReDesign: Recovering Editable Design Structures from Images via Agentic Decomposition

kaist-ai

Submitted by

taesiri

PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space

NanyangTechnologicalUniversity

Nanyang Technological University

Submitted by

geuntaek

Video-Oasis: Rethinking Evaluation of Video Understanding

naver

Submitted by

lhpku20010120

K12-KGraph: A Curriculum-Aligned Knowledge Graph for Benchmarking and Training Educational LLMs

PekingUniversity

Peking University

Submitted by

taesiri

Scaling Mixture-of-Experts Video Pretraining for Embodied Intelligence

robbyant

Submitted by

yangyu90

ResearchStudio-Reel: Automate the Last Mile of Research from Paper to Poster, Video, and Blog

microsoft

Submitted by

fistyyyy

ResearchStudio-Idea: An Evidence-Grounded Research-Ideation Skill Suite from ML Conference Outcomes

microsoft

Submitted by

Ziyang

StateAct: Program State, before Pixels, for Long-Horizon Computer-Use Agents

Salesforce

Salesforce AI Research

2

Submitted by

CNcreator0331

HOMIE: Human-object Centric Video Personalization via Multimodal Intelligent Enchancement

·
11 authors

Submitted by

kpzhang996

AlayaWorld: Interactive Long-Horizon World Modeling -- Full Technical Report

AlayaLab

Submitted by

Rubin-Wei

Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory

·
7 authors

Submitted by

libr6

Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots

SEU-PAISys

PhysicalAI System Group @ SAIL Lab, Southeast University

Submitted by

taesiri

Scalable Visual Pretraining for Language Intelligence

internlm

Intern Large Models

Submitted by

yxl66666

Dual Latent Memory in Vision-Language-Action Models for Robotic Manipulation

·
9 authors

Submitted by

aHapBean

xHC: Expanded Hyper-Connections

dots-studio

Submitted by

taesiri

OvisOCR2 Technical Report

·
13 authors

Submitted by

lhpku20010120

DataPrep-Bench: Benchmarking LLMs as Training Data Preparators

PekingUniversity

Peking University

Submitted by

DogNeverSleep

Beacon: Knowing When and How to Perform Agentic Visual Reasoning

KlingTeam

Submitted by

selimkuzucu

DataComp-VLM: Improved Open Datasets for Vision-Language Models

·
36 authors

Submitted by

LIQIIIII

BadWAM: When World-Action Models Dream Right but Act Wrong

·
3 authors

Submitted by

nstar1125

4D Human-Scene Reconstruction from Low-Overlap Captures

SeoulNatlUniv

Seoul National University

3

Submitted by

ahnGeo

Why Can't I Open My Drawer? Mitigating Object-Driven Shortcuts in Zero-Shot Compositional Action Recognition

naver

Submitted by

taesiri

Cura 1T: Specialized Model for Agentic Healthcare

actava

Submitted by

tianlezeng

ReferTrack: Referring Then Tracking for Embodied Visual Tracking

tencent

Submitted by

hyc2026

FlowMimic: Mask-free Visual Editing and Generation with Pixel-pair Warped Flow Field for Online Video Editing Data Generation and Modality Mimicry

ByteDance-Seed

3

Submitted by

mynameisyoomimi

Scenes as Objects, Not Primitives: Instance-Structured 3D Tokenization from Unposed Views

·
6 authors

2

Submitted by

Bstwpy

BM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation Paradigms

muset-ai

Submitted by

joliang17

Visual Contrastive Self-Distillation

UMCP

University of Maryland College Park

Submitted by

landisen

Morphing into Hybrid Attention Models

ByteDance-Seed

Submitted by

MikaStars39

EvoPolicyGym: Evaluating Autonomous Policy Evolution in Interactive Environments

·
16 authors

Submitted by

WaltonFuture

CLBench-V: Evaluating Multimodal Context Learning from Grounding to Knowledge Acquisition

·
6 authors

Submitted by

Ningyu

LightMem-Ego: Your AI Memory for Everyday Life

zjunlp

2

Submitted by

taesiri

DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation

deepseek-ai

Submitted by

MatanAvitan

MANCE: Manifold Aware Concept Erasure

barilan

Bar-Ilan University