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

Jun 26

Submitted by

taesiri

DanceOPD: On-Policy Generative Field Distillation

ByteDance-Seed

Submitted by

Zuyan

ViQ: Text-Aligned Visual Quantized Representations at Any Resolution

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

Jinyang23

OPID: On-Policy Skill Distillation for Agentic Reinforcement Learning

·
11 authors

Submitted by

taesiri

Qwen-Image-Agent: Bridging the Context Gap in Real-World Image Generation

Qwen

Submitted by

zjj1233

The Verification Horizon: No Silver Bullet for Coding Agent Rewards

Qwen

Submitted by

Snyhlxde

JetSpec: Breaking the Scaling Ceiling of Speculative Decoding with Parallel Tree Drafting

·
12 authors

Submitted by

rebeccazzzz

GUI vs. CLI: Execution Bottlenecks in Screen-Only and Skill-Mediated Computer-Use Agents

·
7 authors

Submitted by

taesiri

Fast LeWorldModel

·
2 authors

Submitted by

jinzhuoran

Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It

CASIA

Chinese Academic of Science Institute of Automation

Submitted by

RunqiLin

Running the Gauntlet: Re-evaluating the Capabilities of Agents Beyond Familiar Environments

Oxford

University of Oxford

Submitted by

Luka-Wang

LISA: Likelihood Score Alignment for Visual-condition Controllable Generation

Submitted by

sinwang

In-Context World Modeling for Robotic Control

OpenMOSS-Team

Submitted by

jaehong31

Confidence-Aware Tool Orchestration for Robust Video Understanding

nanyang-technological-university-singapore

Nanyang Technological University Singapore

Submitted by

taesiri

PhysiFormer: Learning to Simulate Mechanics in World Space

·
3 authors

Submitted by

speed

CoffeeBench: Benchmarking Long-Horizon LLM Agents in Heterogeneous Multi-Agent Economies

·
8 authors

Submitted by

nicklashansen

Hallucination in World Models is Predictable and Preventable

UCSanDiego

University of California at San Diego

Submitted by

viswavi

Discretizing Reward Models

Submitted by

jsonopen1

Information-Aware KV Cache Compression for Long Reasoning

·
4 authors

Submitted by

changdae

Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents

uw-madison

University of Wisconsin - Madison

Submitted by

josefchen

When Does Combining Language Models Help? A Co-Failure Ceiling on Routing, Voting, and Mixture-of-Agents Across 67 Frontier Models

Kaikaku

Submitted by

taesiri

COrigami: An AI Pipeline for Co-Designing Flat-Foldable Visually Recognisable Origami

GoogleDeepMind

Submitted by

Minbyul

OpenBioRQ: Unsolved Biomedical Research Questions for Agents

·
1 authors

Submitted by

hlzhang109

How Post-Training Shapes Biological Reasoning Models

·
8 authors

Submitted by

ll-13

EO-WM: A Physically Informed World Model for Probabilistic Earth Observation Forecasting

·
6 authors

Submitted by

sauradip

ABACUS: Adapting Unified Foundation Model for Bridging Image Count Understanding and Generation

·
3 authors