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

williamium

Self-Supervised Visual On-Policy Distillation

UCSanDiego

University of California at San Diego

Submitted by

WenDingY

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

meituan-longcat

Submitted by

taesiri

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

·
47 authors

Submitted by

wangshengpkucn

Apodex Discovery: Reality Benchmarks and Environments for Evaluating and Building Discoverative Artificial Intelligence

apodex

1

Submitted by

ramiroluo

SimpleOPD: Simple Tokenizer-Agnostic On-Policy Distillation for Long-Context Reasoning

Shanghai-AI-Laboratory

Shanghai AI Laboratory

Submitted by

taesiri

Marionette: Predicting World States, Rendering Geometry, Painting Appearance

·
5 authors

Submitted by

giannor

DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data

SDU-Denmark

University of Southern Denmark (SDU)

Submitted by

Ningyu

MobileMem: Learning from a Year of Mobile Experiences

openkg

OpenKG Consortium

Submitted by

qizekun

HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark

GalaxyGeneralRobotics

Submitted by

TaobaoTmall-AlgorithmProducts

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

alibabagroup

Submitted by

barryhpr

Modular Cognitive Architecture Emerges in Large Language Models

MIT

Massachusetts Institute of Technology

Submitted by

yuheng2000

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

·
25 authors

Submitted by

greeky

Latent On-Policy Self-Distillation

NationalUniversityofSingapore

National University of Singapore

Submitted by

SenXu1123

Claim-Level Reliability Assessment for Efficient Test-Time Reasoning

WeiboAI

Submitted by

zhensuuu

Second Thought: Reasoning in Parallel as LLM Agents Act and Observe

SingaporeManagementUniversity

Singapore Management University

Submitted by

hunarbatra

Multimodal Model Diffing for Feature Discovery and Control

Oxford

University of Oxford

Submitted by

Shuwan

Forecast Collapse in Time-Series Foundation Models

abel-lab

Submitted by

Jana-Z

LittleLearner: Language Models Under Pedagogically Controlled Knowledge Exposure

ELLIS-Institute-Tuebingen

ELLIS Institute Tübingen

Submitted by

haoxintong

Scaling Domain Data Repetition in LLM Pretraining

ByteDance-Seed

Submitted by

JohnLangford

Dion3: Full-Stack Orthogonal Updates

MicrosoftResearch

Microsoft Research

1

Submitted by

sebasmos

Agents Catching Agents: Shortcut Cascades and Benchmark Gaming in Clinical Multi-Agent Systems

MIT

Massachusetts Institute of Technology

Submitted by

SylvainWei

Verifier-Induced Support Reshaping in On-Policy Optimization

PekingUniversity

Peking University

Submitted by

Dvir

UniProbe: A Learnable Token-Level Hallucination Detector for Large VLMs using Multi-Structural Internal Representations

nvidia

Submitted by

jdsouza

A Pathway to General-Purpose Scientific AI: Multimodal Comprehension of Scientific Images

SciKnowOrg

Scientific Knowledge Organization

Submitted by

johnhalloran

Nanbeige4.2-3B on Apple Silicon: Fixing Deployment Bugs and Decreasing Looped Transformer Memory Overhead

·
1 authors

Submitted by

Nyandwi

Amplified Does Not Mean Predictive: Reasoning Behaviors in Thinking Models

CarnegieMellonU

Carnegie Mellon University

Submitted by

emeierkeio

Who Speaks Matters: Authority-Aware Multi-View RAG over Italian Parliamentary Proceedings

unimibb

University of Milan Bicocca

Submitted by

Chidaksh

UNMASK: Discovering and Causally Verifying Spurious Shortcuts in Text Classifiers

UNC-ChapelHill

University of North Carolina at Chapel Hill

2

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