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

April

Submitted by

xxiaoyali

GrandCode: Achieving Grandmaster Level in Competitive Programming via Agentic Reinforcement Learning

deepreinforce-ai

Submitted by

hongyyyyy

Adam's Law: Textual Frequency Law on Large Language Models

FaceMind

Submitted by

zbhpku

DataFlex: A Unified Framework for Data-Centric Dynamic Training of Large Language Models

PekingUniversity

Peking University

Submitted by

737443h

FIPO: Eliciting Deep Reasoning with Future-KL Influenced Policy Optimization

Qwen

Submitted by

tianlezeng

CARLA-Air: Fly Drones Inside a CARLA World -- A Unified Infrastructure for Air-Ground Embodied Intelligence

·
4 authors

Submitted by

jasonrqh

Rethinking Generalization in Reasoning SFT: A Conditional Analysis on Optimization, Data, and Model Capability

AI45Research

Submitted by

taesiri

SkillClaw: Let Skills Evolve Collectively with Agentic Evolver

·
8 authors

Submitted by

jiaruz2

Recursive Multi-Agent Systems

StanfordUniversity

Stanford University

Submitted by

AgPerry

ClawBench: Can AI Agents Complete Everyday Online Tasks?

NAIL-Group

Natural and Artificial Intelligence Lab

Submitted by

taesiri

Tstars-Tryon 1.0: Robust and Realistic Virtual Try-On for Diverse Fashion Items

alibaba-inc

7

Submitted by

weikaih

WildDet3D: Scaling Promptable 3D Detection in the Wild

allenai

Submitted by

taesiri

LLaDA2.0-Uni: Unifying Multimodal Understanding and Generation with Diffusion Large Language Model

inclusionAI

Submitted by

BradyFU

Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding

MME-Benchmarks

Submitted by

taesiri

InCoder-32B-Thinking: Industrial Code World Model for Thinking

·
25 authors

Submitted by

taesiri

Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond

·
42 authors

Submitted by

taesiri

OpenWorldLib: A Unified Codebase and Definition of Advanced World Models

PekingUniversity

Peking University

Submitted by

Zuyan

HY-Embodied-0.5: Embodied Foundation Models for Real-World Agents

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

iieycx

Self-Distilled RLVR

·
10 authors

Submitted by

research4pan

AgentSPEX: An Agent SPecification and EXecution Language

UIUC-ScaleML

UIUC ScaleML Lab

Submitted by

taesiri

Seedance 2.0: Advancing Video Generation for World Complexity

ByteDance-Seed

7

Submitted by

yxl66666

The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlook

·
37 authors

Submitted by

XiaotongLi97

LongCat-Next: Lexicalizing Modalities as Discrete Tokens

meituan-longcat

Submitted by

Musicode

The Past Is Not Past: Memory-Enhanced Dynamic Reward Shaping

·
9 authors

Submitted by

Jawadkotaich

QuanBench+: A Unified Multi-Framework Benchmark for LLM-Based Quantum Code Generation

American University of Beirut

Submitted by

wanderkid

MinerU2.5-Pro: Pushing the Limits of Data-Centric Document Parsing at Scale

opendatalab

4

Submitted by

fredreick

From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company

·
8 authors

Submitted by

taesiri

HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worlds

·
45 authors

Submitted by

yyyang

GameWorld: Towards Standardized and Verifiable Evaluation of Multimodal Game Agents

NationalUniversityofSingapore

National University of Singapore

Submitted by

tobiaslee

Claw-Eval: Toward Trustworthy Evaluation of Autonomous Agents

claw-eval

Submitted by

lhmd

World-R1: Reinforcing 3D Constraints for Text-to-Video Generation

MicrosoftResearch

Microsoft Research

Submitted by

dkliang

When Numbers Speak: Aligning Textual Numerals and Visual Instances in Text-to-Video Diffusion Models

H-EmbodVis

Submitted by

AaronHuangWei

TriAttention: Efficient Long Reasoning with Trigonometric KV Compression

nvidia

Submitted by

wenyi

GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents

zai-org

Submitted by

hbx

Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe

TsinghuaNLP

Tsinghua NLP Group

Submitted by

JasperHaozhe

RationalRewards: Reasoning Rewards Scale Visual Generation Both Training and Test Time

TIGER-Lab

Submitted by

wangzx1994

Generative World Renderer

ShandaAI

Shanda AI Research Tokyo

Submitted by

HasuerYu

KnowRL: Boosting LLM Reasoning via Reinforcement Learning with Minimal-Sufficient Knowledge Guidance

TianjinUniversity

Tianjin University

Submitted by

Gaojunyao

MegaStyle: Constructing Diverse and Scalable Style Dataset via Consistent Text-to-Image Style Mapping

tencent

Submitted by

LZXzju

SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization

·
10 authors

Submitted by

patricebechard

Terminal Agents Suffice for Enterprise Automation

ServiceNow-AI

7

Submitted by

ldkong

OneVL: One-Step Latent Reasoning and Planning with Vision-Language Explanation

xiaomi-research

Xiaomi Research

Submitted by

HideOnBush

FORGE:Fine-grained Multimodal Evaluation for Manufacturing Scenarios

UWaterloo

University of Waterloo

Submitted by

chengtan9907

Programming with Data: Test-Driven Data Engineering for Self-Improving LLMs from Raw Corpora

opendatalab

Submitted by

XuGuo699

CoInteract: Physically-Consistent Human-Object Interaction Video Synthesis via Spatially-Structured Co-Generation

alibaba-inc

Submitted by

dongguanting

Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence

ByteDance-Seed

4

Submitted by

yhx12

GEMS: Agent-Native Multimodal Generation with Memory and Skills

·
7 authors

Submitted by

zunhai

Attention Sink in Transformers: A Survey on Utilization, Interpretation, and Mitigation

meituan-longcat

Submitted by

chanyoungkim

LIBERO-Para: A Diagnostic Benchmark and Metrics for Paraphrase Robustness in VLA Models

HAI-Lab

Human-centered AI Laboratory

Submitted by

csuhan

OpenGame: Open Agentic Coding for Games

·
11 authors

Submitted by

alkxncda

LPM 1.0: Video-based Character Performance Model

·
25 authors

Submitted by

iieycx

Near-Future Policy Optimization

·
9 authors

Submitted by

bibona

Lingshu-Cell: A generative cellular world model for transcriptome modeling toward virtual cells

Alibaba-DAMO-Academy

8

Submitted by

Lines

Large Language Models Explore by Latent Distilling

ShanghaiTech

ShanghaiTech University

Submitted by

Jaafar

RADIO-ViPE: Online Tightly Coupled Multi-Modal Fusion for Open-Vocabulary Semantic SLAM in Dynamic Environments

BE2R

Submitted by

play0718

A Simple Baseline for Streaming Video Understanding

lmms-lab

Submitted by

donghao-zhou

OmniShow: Unifying Multimodal Conditions for Human-Object Interaction Video Generation

ByteDance

Submitted by

leiiizhang

Think in Strokes, Not Pixels: Process-Driven Image Generation via Interleaved Reasoning

Submitted by

taesiri

Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation

AI-at-Meta

Submitted by

taesiri

EXAONE 4.5 Technical Report

LGAI-EXAONE

Submitted by

KID-22

Learning to Retrieve from Agent Trajectories

RUC-GSAI-IIRLab

RUC-GSAI-IIRLab

Submitted by

Fangda-Ye

MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcome

miromind-ai

Submitted by

zd294cam

Project Imaging-X: A Survey of 1000+ Open-Access Medical Imaging Datasets for Foundation Model Development

General-Medical-AI

General Medical AI

Submitted by

xytian1008

All Roads Lead to Rome: Incentivizing Divergent Thinking in Vision-Language Models

Submitted by

eamonn-zh

ReVSI: Rebuilding Visual Spatial Intelligence Evaluation for Accurate Assessment of VLM 3D Reasoning

SFU

Simon Fraser University

Submitted by

ZihanWang314

RAGEN-2: Reasoning Collapse in Agentic RL

MLL-Lab

Submitted by

gregH

OccuBench: Evaluating AI Agents on Real-World Professional Tasks via Language World Models

Qwen

Submitted by

ZhaochongAn

VGGRPO: Towards World-Consistent Video Generation with 4D Latent Reward

google

3

Submitted by

SagiPolaczek

Video Analysis and Generation via a Semantic Progress Function

tau

Tel Aviv University

Submitted by

taesiri

SpatialEvo: Self-Evolving Spatial Intelligence via Deterministic Geometric Environments

·
19 authors

Submitted by

FuNz

DiPO: Disentangled Perplexity Policy Optimization for Fine-grained Exploration-Exploitation Trade-Off

ECNU

East China Normal University

2

Submitted by

taesiri

Qwen3.5-Omni Technical Report

·
1 authors

Submitted by

Kimhi

Maximal Brain Damage Without Data or Optimization: Disrupting Neural Networks via Sign-Bit Flips

nvidia

Submitted by

taesiri

Memory Intelligence Agent

·
9 authors

Submitted by

JonaRuthardt

Steerable Visual Representations

FunAILab

Fundamental AI Lab at UTN

Submitted by

huaijinpi

Strips as Tokens: Artist Mesh Generation with Native UV Segmentation

DEEMOSTECH

DEEMOS Technology

Submitted by

quao627

CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery

Massachusetts-Institute-of-Technology

Massachusetts Institute of Technology

Submitted by

chengtim

VOID: Video Object and Interaction Deletion

netflix

Submitted by

taesiri

Embarrassingly Simple Self-Distillation Improves Code Generation

apple

Submitted by

KID-22

DR-Venus: Towards Frontier Edge-Scale Deep Research Agents with Only 10K Open Data

inclusionAI

Submitted by

sun0o0

ACES: Who Tests the Tests? Leave-One-Out AUC Consistency for Code Generation

·
7 authors

4

Submitted by

Zigeng

DMax: Aggressive Parallel Decoding for dLLMs

NationalUniversityofSingapore

National University of Singapore

Submitted by

taesiri

Matrix-Game 3.0: Real-Time and Streaming Interactive World Model with Long-Horizon Memory

·
23 authors

Submitted by

Chenyu-Zhou

Externalization in LLM Agents: A Unified Review of Memory, Skills, Protocols and Harness Engineering

SJTU

Shanghai Jiao Tong University

2

Submitted by

jinpeng0528

AURA: Always-On Understanding and Real-Time Assistance via Video Streams

·
12 authors

Submitted by

taesiri

ClawGym: A Scalable Framework for Building Effective Claw Agents

·
13 authors

4

Submitted by

taesiri

OpenVLThinkerV2: A Generalist Multimodal Reasoning Model for Multi-domain Visual Tasks

uclanlp

Submitted by

vinthony

CutClaw: Agentic Hours-Long Video Editing via Music Synchronization

GVCLab

GVC Lab at Great Bay University

Submitted by

N2048M

Turning the TIDE: Cross-Architecture Distillation for Diffusion Large Language Models

PekingUniversity

Peking University

Submitted by

LZXzju

KnowU-Bench: Towards Interactive, Proactive, and Personalized Mobile Agent Evaluation

·
16 authors

Submitted by

Tsumugii

GBQA: A Game Benchmark for Evaluating LLMs as Quality Assurance Engineers

·
3 authors

Submitted by

taesiri

SmartPhotoCrafter: Unified Reasoning, Generation and Optimization for Automatic Photographic Image Editing

·
13 authors

Submitted by

Haoyuwu

MultiWorld: Scalable Multi-Agent Multi-View Video World Models

hkuhk

The University of Hong Kong

Submitted by

Tyrannosaurus

MegaTrain: Full Precision Training of 100B+ Parameter Large Language Models on a Single GPU

·
4 authors

Submitted by

HyunsooCha

Vanast: Virtual Try-On with Human Image Animation via Synthetic Triplet Supervision

SeoulNatlUniv

Seoul National University

Submitted by

csfufu

Unify-Agent: A Unified Multimodal Agent for World-Grounded Image Synthesis

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

LIQIIIII

Vision-Language-Action Safety: Threats, Challenges, Evaluations, and Mechanisms

·
9 authors

Submitted by

taesiri

MolmoWeb: Open Visual Web Agent and Open Data for the Open Web

·
16 authors

Submitted by

gong97

Uni-ViGU: Towards Unified Video Generation and Understanding via A Diffusion-Based Video Generator

·
9 authors

Submitted by

limuloo1999

RefineAnything: Multimodal Region-Specific Refinement for Perfect Local Details

Zhejiang University

Submitted by

groundhogLLM

Beyond Accuracy: Unveiling Inefficiency Patterns in Tool-Integrated Reasoning

ustc-community

University of Science and Technology of China

Submitted by

Shenzhi

Can LLMs Learn to Reason Robustly under Noisy Supervision?

·
10 authors

Submitted by

he-zh22

Vision2Web: A Hierarchical Benchmark for Visual Website Development with Agent Verification

zai-org

Submitted by

VincentHancoder

ViGoR-Bench: How Far Are Visual Generative Models From Zero-Shot Visual Reasoners?

meituan

Submitted by

Zeyue7

Audio-Omni: Extending Multi-modal Understanding to Versatile Audio Generation and Editing

·
11 authors

Submitted by

taesiri

DV-World: Benchmarking Data Visualization Agents in Real-World Scenarios

·
20 authors