每日论文 Daily Papers

来自 Hugging Face 与研究社区 · 中英对照

Daily Weekly Monthly

‹ 2026-08-03 ~ 2026-08-09 ›

跳转日期： 11-01 39 10-28 35 10-27 31 10-24 50 10-23 52 10-22 60 10-21 64 10-20 64 10-17 83 10-16 74 10-15 99 10-14 22 10-13 58 10-11 4 10-10 68 10-09 61 10-07 72 10-03 82 10-02 84 10-01 86 09-30 95 09-29 31 09-27 22 09-25 29 09-24 23 09-23 30 09-22 30 09-20 1 09-19 23 09-18 23

LongHorizon-Harness: Advancing Long-Horizon Agents for Real-World Tasks

▲ 115 💬 2 ★ 154

SwanTale: Unified Multi-Speaker Speech and Audio Generation for Instruct and Zero-Shot Tasks

▲ 111 💬 7

MerchantBench: Benchmarking LLM Agents for Long-Term Coherence in E-Commerce Operations

▲ 82 💬 3 ★ 29

JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion

▲ 75 ★ 138

AURORA-LM: Autoencoding Unified Representation for Continuous-Latent Diffusion Language Modeling

▲ 71 💬 1 ★ 21

Hunyuan3D-Buffalo 1.0: A Unified Multimodal Model for Scalable 3D Generation, Understanding, and Editing

▲ 69 💬 1

From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement

▲ 68 💬 2 ★ 138

Mental World Modeling

▲ 52 💬 1 ★ 1

ABSeeker: Training Long-Horizon Search Agents via Answer-Backtracked Credit Assignment

▲ 52 💬 1

N_0-VTLA: Scaling Vision-Tactile-Language-Action Model with Latent Tactile Tokens

▲ 49 💬 1 ★ 36

AgentOPSD: Recursive Self-Distillation for Agentic Reinforcement Learning

▲ 47 💬 1 ★ 2

DAPD: Dual-Anchored Policy Distillation

▲ 46 💬 2

ToolArtist: Tool-Using Unified Multimodal Models for Agentic Image Generation

▲ 45

Progressive Agent Skill Generation via Reinforcement Learning

▲ 44 💬 1 ★ 15

VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillation

▲ 43 💬 1 ★ 34

Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent

▲ 43 💬 1

Meshy T2: Fast Native Mesh Generation with Flow Matching

▲ 40

Knowledge-Geometry Decoupling: Refreshable Pretrained Transfer for Streaming Recommendation

▲ 38 💬 1 ★ 3

Towards Physics of Multimodal Pretraining: Knowledge Flow, Modality Synergy, Early Unification, and Recipes

▲ 38 💬 1

PCSD: Persistent Consistency for Self-Distillation in Agentic Reinforcement Learning

▲ 37 💬 1

The Personalization Mirage: How LLMs Fabricate User Profiles, and Why Self-Monitoring Misleads

▲ 37 💬 2

WorldClaw: Agentic 3D Open-World Generation at Scale

▲ 35

WorldExam: Benchmarking World Models from Apparent Appearance to Inherent Reactivity

▲ 30 💬 2 ★ 12

AISPA: User-Centric System Prompt Auditing for Large Language Model Applications

▲ 29 💬 1 ★ 5

UEmbed: Unified Sparse and Dense Multimodal Embeddings

▲ 29 💬 1 ★ 7

CADENA: Stepwise CAD Reverse Engineering

▲ 29 💬 1 ★ 13

PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents

▲ 28 ★ 6

Quo Vadis, World Modeling?

▲ 28 💬 1 ★ 26

Deferred Exposure of Future Trajectories for Verifiable Reasoning in Autonomous Driving VLMs

▲ 27 💬 1 ★ 20

OneDayAgent: Towards a Long-Horizon Harness for Autonomous Agents

▲ 27 💬 2 ★ 7

N_0-TWAM: Scaling Tactile-Native World-Action Model for Contact-Rich Manipulation

▲ 26 💬 1 ★ 48

SKT: Skill-Use Training at Scale via Verified Synthetic Data Generation

▲ 25 💬 1

Learning from Failures: Retrieval-Centric CoT via Hard Negatives for Unified Multimodal Retrieval

▲ 25 💬 1 ★ 3

EnvACE: Internalizing Environment Dynamics via World Rehearsal for Agentic Reinforcement Learning

▲ 25 💬 1 ★ 3

ChronoVision: Temporal Reasoning via Latent State Reconstruction

▲ 24 💬 1

Scaling Properties of Text Conditioning in Visual Generation

▲ 23 💬 1 ★ 55

QQWorld: Quantile-Quantile Matching for World Model Regularization

▲ 23 💬 1

OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models

▲ 23 💬 1 ★ 6

SAF-OPD: Stable Advantage Fusion for On-Policy Distillation

▲ 22 💬 1

SWE-Touch: Benchmarking Coding Agents When Users Touch the Code

▲ 22 💬 1 ★ 4

WCM: A World Critic Model for Vision-Language-Action Reinforcement Learning

▲ 22 💬 6 ★ 17

LLaDA MoE v2: Scaling Mixture-of-Experts Diffusion Language Models

▲ 22 💬 1

OmniPack: Unified Token Compression for Efficient Omni-modal Large Language Models

▲ 21 💬 1 ★ 4

Any-OPD: Heterogeneous On-Policy Distillation for Flow-Matching Models via Representation-Space Bridging

▲ 21 💬 2

GDPevo: Evaluating Agent Self-Evolution on Real Business Tasks

▲ 21 💬 1 ★ 48

GradCuit: Credit-Assigned Gradient Flow Enables Robust and Interpretable Test-Time Latent Reasoning

▲ 20 💬 2 ★ 3

Motion Beyond Morphology: Bootstrapping Cross-Category Motion Transfer from Abstract Motion Representations

▲ 20 💬 3

CAPEval: A Decoupled Caption Evaluation across Understanding and Generation

▲ 20 💬 1 ★ 9

Toward Skill-Native LLMs: Skill Entropy for Benchmarking and Training Long-Horizon Reasoning

▲ 20 💬 1 ★ 4

Fewer Clarifications, Better Code: Benchmarking Cross-Session Personalized Ambiguity Adaptation in Coding Assistants

▲ 19 💬 2

DiffusionGemma Technical Report

▲ 19

SkillJack: Persistent Skill Backdoors in Self-Evolving Agents

▲ 19 💬 1

UniWorld-Design: From Pixel Generation to Layer-Native Design

▲ 19 💬 1

When Teachers Mislead: Spurious-Signal-Aware On-Policy Distillation

▲ 19 💬 1 ★ 1

On-Policy Delta Distillation for Multilingual Math Reasoning

▲ 18 💬 1

To Add Is Machine, To Delete Is Human: Measuring and Mitigating Deletion Avoidance in LLM Code Editing

▲ 17 💬 1

GPTQ-2D: Cubic-Time Two-Sided Adaptive Rounding

▲ 17 💬 1

TurnSight: Turn-Level Hindsight Self-Distillation for Tool-Integrated Reasoning

▲ 17 💬 1 ★ 2

NOLLI: A Difficulty-Calibrated Puzzle Benchmark for Diagnosing the English-Korean Performance Gap

▲ 17 💬 3 ★ 4

Ego2Robot: Scalable Robot Data Synthesis from Egocentric Human Data

▲ 17 💬 1

HarnessOpt-Bench: Evaluating LLMs at Harness Optimization

▲ 17 💬 1

Roomer: Reflective Object-Grounded Model Editing and Repair for 3D Indoor Layout Synthesis

▲ 16 💬 1

GST-Bench: Can VLMs Develop Global Spatial Awareness from Video?

▲ 16

Enhancing Rubric-based RL via Self-Distillation

▲ 15 💬 1

DeepVoyager-VL: Incentivizing Vision-in-the-Loop Search for Long-Horizon Multimodal Agents

▲ 15 💬 1 ★ 1

AVE-Compass: Towards Holistic Evaluation for Audio-Video Editing Abilities

▲ 15 💬 1 ★ 5

3DZip: Spatial-Aware Feature Diversity-Guided Token Compression for 3D Question Answering

▲ 14 💬 2 ★ 13

GROVE: Growing and Reasoning over Temporally Stratified Memory from Streaming Video Experience

▲ 14 💬 1

DyPES-VLA: Learning Shared Dynamics Priors and Embodiment-Specific Control for Cross-Embodiment Manipulation

▲ 14

World-to-Wrist: Task-Conditioned Future Wrist Modeling for Fine-Grained Robot Manipulation

▲ 14 💬 1 ★ 9

ExtractBench: A Benchmark for Schema-Guided Enterprise Document Extraction

▲ 13

LeapTalk: Breaking the Latency-Quality Trade-off in Talking Head Generation

▲ 13 💬 1 ★ 12

StyleForge: Indoor Furniture Styling by Counterfactual Reasoning in a Hypergraph Field

▲ 12 💬 1

DreamTraj: Generating 6-DoF Object Trajectories by Reading Unrendered Video Diffusion Latents

▲ 12 💬 1

ExplainBench: Evaluating Code Explanations from Agents

▲ 12 💬 2

K-EXAONE 2.0 Technical Report

▲ 12

When Memory Lies: An Empirical Study of Spatial Memory Staleness in VLM Agents

▲ 12 💬 1

Distill Where You Fail: Recovering Learning Signals of Negative RL-Groups from Adaptive Teacher Guidance

▲ 12 💬 1

Evaluation-Verification Reward for Consistent Multi-Reference Image Editing

▲ 11

ContinualSkillBench: Can LLM Agents Truly Evolve Their Capabilities?

▲ 11 💬 3 ★ 2

HelloWorld: Enabling Socially Interactive Characters in Video World Models

▲ 11 ★ 9

Poplar: A Scalable Pipeline for Human-Centric Image Dataset Synthesis

▲ 10 💬 3 ★ 5

FocusMem: Factorizing Content, Readout, and Trust in Latent GUI Memory

▲ 10 💬 1

CalibForge: Adversarial Solver Calibration for Scaling Learnable Terminal Tasks

▲ 10 💬 1 ★ 4

ScrambleToolBench: Agents Search Exhaustively Even When Their Own Map Points to the Next Step

▲ 9 💬 1

MiniWorld: Democratizing the Training of Video World Models from Scratch

▲ 9 💬 1 ★ 21

Decoding Children's Gait Behavior

▲ 9 💬 1 ★ 2

Are the Financial Reasoning from LLMs Credible? A Real World Test over Long-Horizon Statements

▲ 9 💬 1

SKILL-KD: Contrastive Skill Distillation for LLM Agents

▲ 9 💬 1

When Agents Learn to Be You: Benchmarking Privacy Leakage, Impersonation Risk, and Defenses in Persona Skills

▲ 8 💬 1 ★ 1

WorldCycle: Self-Verifiable Reinforcement Learning for Long-Horizon Video World Models

▲ 8 💬 1

OPD-V: Visual On-Policy Self-Distillation with Modality Balance

▲ 8

SmartMage: Dynamic Modality Orchestration for 3D Scene Understanding

▲ 8 💬 1

ODEWorld: A Continuous Predictive Architecture via Physical-Time Flow

▲ 7 💬 1 ★ 11

RecHarness: A Bandit-Routed Agentic Harness for Self-Evolving Recommender Systems

▲ 7 💬 1 ★ 1

Model or Harness? An Interaction-Centric Taxonomy for Localizing Agent Failures

▲ 7 💬 1

BridgeVLA++: A Data-Efficient, Generalizable, and Memory-Augmented Vision-Language-Action Framework for 3D Manipulation

▲ 7 💬 1

Poly-OPD: Heterogeneous Multi-Teacher On-Policy Distillation for Capability-Selectable Flow Models

▲ 7 💬 2

Lossless Tensor Compression as Program Synthesis

▲ 7 💬 1 ★ 1

PaDoc: Layout-Grounded Parallel Decoding for Document Parsing

▲ 7 💬 1 ★ 2

Relax Within, Balance Across: Geometry-Guided Load Balancing for Vision-Language Mixture-of-Experts

▲ 6 💬 1 ★ 1

Push-Wiper: Toward General-Purpose Robotic Cleaning across Varied Stains and Surfaces with Segmented Pushing Trajectories

▲ 6 💬 1

ST-WAM: Semantic-Temporal World Action Model for Robust Manipulation under Visual Distribution Shifts

▲ 6 💬 1 ★ 8

UniWorld-View: Large-Baseline View Synthesis via Video Diffusion Models

▲ 6 💬 1 ★ 100

RestoreKV: Recovering Full-Cache Behavior Under Aggressive Query-Agnostic KV Cache Eviction

▲ 5 💬 1

Agent Against Agent: An Agentic System for Automatic Prompt Injection Red Teaming

▲ 5 💬 1 ★ 2

Consistency-Driven Co-Evolution for Self-Supervised Cross-Representation Learning

▲ 5 💬 1

TriGlue: a Biology-Inspired Generative Model for Generating Molecular Glue-Induced Ternary Complex

▲ 5 💬 1 ★ 2

EMBL AI Librarian: Life-Sciences Knowledge Layer for AI Agents

▲ 4 💬 1 ★ 8

RL^2-VLA: Adaptive RL Latent Compositional Steering with Test-Time Scaling for Vision-Language-Action Models

▲ 4 💬 2 ★ 8

ICDAR 2026 Competition on Information Extraction from Atomic Layer Deposition/Etching (ALD/E) Scientific Figures

▲ 4 💬 1 ★ 4

When Attention Goes Blind: Numerical Failure in ALiBi Positional Encodings

▲ 4 💬 1

PosterMELD: Multi-Agent Paper-to-Poster Generation for Controllable Design Diversity with Editable Print-Ready Outputs

▲ 4 💬 1 ★ 1

One Future, Every Robot: Label-Efficient Collective-State Prediction with Decentralized JEPA

▲ 3 💬 1

Loud or Silent? A Reusable Framework for Per-Modality Failure Analysis in Multimodal Clinical AI

▲ 3 💬 1 ★ 5

Zero-Mem: Zero-Token Memory Operations for LLM Agents

▲ 3 💬 1

Compute Globally, Materialize Locally: The Memory Contract of Sparse Event-KV

▲ 3 💬 1 ★ 2

Multi-Task Multi-Frame Visual Piano Transcription

▲ 3 💬 1 ★ 3

ARCHead: Activation-Metric Residual Correction for Large Language Model Output Heads

▲ 3 💬 1 ★ 3

Better, Stronger, Faster, and Broader: Structured All-Mask Prediction for MLLM-Based Segmentation

▲ 3 💬 1 ★ 40

Self-Evolving Coding Agents

▲ 3 💬 1 ★ 7

ContextMaster: Interactive Multi-Shot Video Creation via Fixed-Budget Sparse Context Routing

▲ 3 💬 1

Toward Robust and 3D-Aware RGB-NIR Imaging in the Dark

▲ 2 💬 3 ★ 1

SULAND v2: A Refined RGB Dataset and Deep Learning Object Detection Benchmark for UAV/UGV-Based SUrface LANDmine Detection Under Domain Shift

▲ 2 💬 1 ★ 2

Beyond Feeling Better: Capability-Sustaining Emotional Dialogue as a Longitudinal Research Paradigm

▲ 2 💬 1

Not All Tokens Deserve Equal Credit: Counterfactual Sensitivity Credit Reallocation for Long-CoT Reasoning

▲ 2 💬 1

Wnuan: Staged Post-Training for Question Answering over Proprietary Enterprise Knowledge

▲ 2 💬 1

When Many Answers Are Valid, Voting Fails: Symbolic Verification for Best-of-K Causal Reasoning in LLMs

▲ 2 💬 1

LegalPincite: Multi-level Legal Information Retrieval Dataset

▲ 2 💬 1

SIGNPOST-Bench: Benchmarking Text-Vision Conflict Resolution in Multimodal Large Language Models

▲ 2 💬 1

DRIFT: Derailing Denoising Trajectories of Flow-Matching VLAs with Adversarial Patch Attack

▲ 2 💬 1

EffectLearner: World-Aware Object-Effect Reasoning for Real-World Video Object Removal

▲ 2 💬 2

Invisible Shortcuts: Why Vision Encoders Know Your Camera

▲ 2 💬 1

Safeguards Based on Copyable Context Cannot Provide Reliable Safety for LLMs

▲ 1 💬 1

In the Driver's Seat: A Multi-Company Study on the Reality of Autonomous Driving System Testing

▲ 1 💬 1

GEOID-Flood: A Large-Scale Multi-Modal Benchmark Dataset for Flood Segmentation

▲ 1 💬 1 ★ 1

SG-WAM: Self-Guided World Modeling in Geometry-Aware Policy Space

▲ 1 💬 1

A Frozen Pixel-Space Diffusion Model Can Guide Itself with Its Own Samples

▲ 1 💬 1 ★ 2

Seeing or Knowing? Visual Context Sensitivity in Multimodal Large Language Models

▲ 1 💬 1

ReflectRL: Learning from Golden Negative Trajectories via Reflective-to-Direct Reasoning

▲ 1 💬 1 ★ 1

ChronoLens: Measuring Language Change Across Time, Languages, and Linguistic Levels

▲ 1 💬 1

Resume Means Resume: A Machine-Checked Conformance Contract for Checkpoint, Interrupt, and Resume Semantics in Workflow Persistence Layers

▲ 1 💬 1

Would You Walk to the Car Wash? Revealing the Salience Bias of Large Language Models in Commonsense Reasoning

▲ 0 💬 1

SGTP: Sampling-based Game-Theoretic Planning for Real-Time Multi-Vehicle Autonomous Racing

▲ 0 💬 1 ★ 3

CURV: Enhancing Chart Understanding Through Curriculum Visual Grounded Reasoning

▲ 0 💬 1