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

Aug 17

Submitted by

heisejiasuo

Can We Defend Against AI-Generated Video Attacks on Real-World Crisis Events? A Systematic Evaluation of Detectors, Generators and Social Dissemination

huggingface

Submitted by

taesiri

Intern-S2-Mobius: Foundation Model with Decoupled Knowledge and Reasoning

·
47 authors

Submitted by

WenDingY

Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and Development

meituan-longcat

Submitted by

qizekun

HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark

GalaxyGeneralRobotics

Submitted by

Ningyu

MobileMem: Learning from a Year of Mobile Experiences

openkg

OpenKG Consortium

Submitted by

TaobaoTmall-AlgorithmProducts

CPI-Bench: A Comprehensive,Practical and Intelligent Benchmark for Real-World Image Editing

alibabagroup

Submitted by

SenXu1123

Claim-Level Reliability Assessment for Efficient Test-Time Reasoning

WeiboAI

Submitted by

taesiri

Marionette: Predicting World States, Rendering Geometry, Painting Appearance

·
5 authors

Submitted by

yuheng2000

PRM-as-a-Judge 1.5: A Toolkit for Robot Process Assessment

·
25 authors

Submitted by

haoxintong

Scaling Domain Data Repetition in LLM Pretraining

ByteDance-Seed

Submitted by

greeky

Latent On-Policy Self-Distillation

NationalUniversityofSingapore

National University of Singapore

Submitted by

hunarbatra

Multimodal Model Diffing for Feature Discovery and Control

Oxford

University of Oxford

Submitted by

JohnLangford

Dion3: Full-Stack Orthogonal Updates

MicrosoftResearch

Microsoft Research

1

Submitted by

SylvainWei

Verifier-Induced Support Reshaping in On-Policy Optimization

PekingUniversity

Peking University

Submitted by

taesiri

SPARGen: Unifying Spatial Perception and Reasoning through Native Multimodal Generation

·
10 authors

Submitted by

IceInPot

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction

ByteDance