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

Jan 19

Submitted by

Yikunb

Your Group-Relative Advantage Is Biased

·
13 authors

5

Submitted by

EilamSha

The Poisoned Apple Effect: Strategic Manipulation of Mediated Markets via Technology Expansion of AI Agents

Technion

Technion Israel institute of technology

2

Submitted by

naiweizi

Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text

meituan-longcat

2

Submitted by

vzl123

RubricHub: A Comprehensive and Highly Discriminative Rubric Dataset via Automated Coarse-to-Fine Generation

·
9 authors

Submitted by

Jeryi

When Personalization Misleads: Understanding and Mitigating Hallucinations in Personalized LLMs

RUC

Renmin University of China

Submitted by

sundrops

ACoT-VLA: Action Chain-of-Thought for Vision-Language-Action Models

agibot-world

2

Submitted by

taesiri

BAPO: Boundary-Aware Policy Optimization for Reliable Agentic Search

·
10 authors

Submitted by

lucianodelcorro

Entropy Sentinel: Continuous LLM Accuracy Monitoring from Decoding Entropy Traces in STEM

UdeSA

Universidad de San Andrés

2

Submitted by

taesiri

FrankenMotion: Part-level Human Motion Generation and Composition

·
5 authors

Submitted by

taki555

ProFit: Leveraging High-Value Signals in SFT via Probability-Guided Token Selection

IIGroup

Tsinghua IIGroup

Submitted by

yawarnihal

ShapeR: Robust Conditional 3D Shape Generation from Casual Captures

Submitted by

taesiri

Reasoning Models Generate Societies of Thought

·
5 authors

Submitted by

user0102

PersonalAlign: Hierarchical Implicit Intent Alignment for Personalized GUI Agent with Long-Term User-Centric Records

·
5 authors

3

Submitted by

taesiri

Building Production-Ready Probes For Gemini

·
7 authors

Submitted by

BiaoGong

PhysRVG: Physics-Aware Unified Reinforcement Learning for Video Generative Models

·
10 authors

2

Submitted by

rajkumarrawal

AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts

GAIR

Submitted by

adrianb1

More Images, More Problems? A Controlled Analysis of VLM Failure Modes

SamsungResearch

Samsung Research

Submitted by

taesiri

AstroReason-Bench: Evaluating Unified Agentic Planning across Heterogeneous Space Planning Problems

·
5 authors

Submitted by

beiweixiaoxu

Language of Thought Shapes Output Diversity in Large Language Models

iNLP-Lab

iNLP Lab @ SUTD

Submitted by

rajkumarrawal

What Matters in Data Curation for Multimodal Reasoning? Insights from the DCVLR Challenge

·
8 authors

Submitted by

zjuyb

PhyRPR: Training-Free Physics-Constrained Video Generation

·
4 authors

2

Submitted by

kahnchana

Future Optical Flow Prediction Improves Robot Control & Video Generation

Salesforce