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

May 26

Submitted by

Nothing2Say

DVAO: Dynamic Variance-adaptive Advantage Optimization for Multi-reward Reinforcement Learning

·
6 authors

2

Submitted by

Kaining

WBench: A Comprehensive Multi-turn Benchmark for Interactive Video World Model Evaluation

meituan-longcat

Submitted by

anchen1011

Macaron-A2UI: A Model for Generative UI in Personal Agents

mindlab-research

4

Submitted by

Bang-UdeM-Mila

Foundation Protocol: A Coordination Layer for Agentic Society

University of Montreal

Submitted by

lhmd

TriSplat: Simulation-Ready Feed-Forward 3D Scene Reconstruction

zju

Zhejiang University

Submitted by

HansonDJN

Toward Native Multimodal Modeling: A Roadmap

tencent

Submitted by

Yeonjun

Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents

Submitted by

mwxely

ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning

lmms-lab

Submitted by

hsaest

QUEST: Training Frontier Deep Research Agents with Fully Synthetic Tasks

osunlp

Submitted by

joesharratt29

ThriftAttention: Selective Mixed Precision for Long-Context FP4 Attention

·
1 authors

Submitted by

tgy2024

AutoResearch AI: Towards AI-Powered Research Automation for Scientific Discovery

·
23 authors

Submitted by

HanSolo9682

Your Embedding Model is SMARTer Than You Think

·
6 authors

Submitted by

Haiyang-W

Claw-Anything: Benchmarking Always-On Personal Assistants with Broader Access to User's Digital World

·
11 authors

Submitted by

BryanWangNLP

CUA-Gym: Scaling Verifiable Training Environments and Tasks for Computer-Use Agents

Qwen

Submitted by

Koi953215

Pantheon360: Taming Digital Twin Generation via 3D-Aware 360° Video Diffusion

BoschUS

Submitted by

yang29

On-Policy Adversarial Flow Distillation for Autoregressive Video Generation

·
7 authors

2

Submitted by

dericky286

ControlLight: Towards Controllable, Consistent, and Generalizable Low-Light Enhancement

·
7 authors

Submitted by

donghao-zhou

SkillEvolBench: Benchmarking the Evolution from Episodic Experience to Procedural Skills

ohiostate

The Ohio State University

Submitted by

jhhuangchloe

Recursive Flow Matching

·
4 authors

1

Submitted by

Alex7616

Anticipate and Learn: Unleashing Idle-Time Compute in Proactive Agents

SJTU

Shanghai Jiao Tong University

Submitted by

Songweii

Channel-wise Vector Quantization

·
8 authors

Submitted by

Concyclics

MemForest: An Efficient Agent Memory System with Hierarchical Temporal Indexing

·
9 authors

Submitted by

CircleRadon

InstructSAM: Segment Any Instance with Any Instructions

·
9 authors

Submitted by

domejiraphon

Helix4D: Complex 4D Mesh Generation

snap-research

2

Submitted by

yoavgur

Faithfulness Metrics Don't Measure Faithfulness: A Meta-Evaluation with Ground Truth

tau

Tel Aviv University

Submitted by

isno0907

Geometry-Aware Image Flow Matching

·
3 authors

Submitted by

smcleish

Language Models Need Sleep

·
4 authors

3

Submitted by

Sanae-Kochiya-2003

CoSPlay: Cooperative Self-Play at Test-Time with Self-Generated Code and Unit Test

sanae-ai-lab

Submitted by

lilvjosephtang

RankJudge: A Multi-Turn LLM-as-a-Judge Synthetic Benchmark Generator

Layer6

Submitted by

Tangc03

Towards Customized Multimodal Role-Play

·
8 authors

Submitted by

HU22333

SEAL: Synergistic Co-Evolution of Agents and Learning Environments

WestlakeUniversity

Westlake University

Submitted by

Arturjssln

CRONOS: Benchmarking Counterfactual Physical Consistency in Video Models

·
3 authors

Submitted by

lzq2021

MetaphorVU: Towards Metaphorical Video Understanding

·
16 authors

Submitted by

vshrivas

ECHO: Terminal Agents Learn World Models for Free

MicrosoftResearch

Microsoft Research

Submitted by

wafer-bob

Coloring the Noise: Adversarial Sobolev Alignment for Faithful Image Super Resolution

CASIA

Chinese Academic of Science Institute of Automation

Submitted by

flying-iwik

PRISM: Position-encoded Regressive Inverse Spectral Model for Multilayer Thin-Film Design

·
4 authors

Submitted by

iasonsky

Injecting Image Guidance into Text-Conditioned Diffusion Models at Inference

uva

University of Amsterdam

Submitted by

zhiminy

Towards Evaluation Engineering: An Empirical Study of ML Evaluation Harnesses in the Wild

Queen's University

Submitted by

ZQ-Dev

How Far Will They Go? Red-Teaming Online Influence with Large Language Models

usc-isi

USC Information Sciences Institute

Submitted by

Harahan

Reinforcing Few-step Generators via Reward-Tilted Distribution Matching

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

NicolasCC

HorizonStream: Long-Horizon Attention for Streaming 3D Reconstruction

Hong Kong University of Science and Technology(GuangZhou)

Submitted by

hsi1032

Cross-scale Aligned Supervision for Training GANs

·
3 authors

1

Submitted by

hongst

SemBridge: Language Transfer in Sparse Encoders via Multilingual Semantic Bridges

nlpai-lab

NLP & AI - Korea University

Submitted by

dwenlong

Directional Alignment Mitigates Reward Hacking in Reinforcement Learning for Language Models

·
7 authors

Submitted by

Zacharyvixx

SimuWoB: Simulating Real-World Mobile Apps for Fast and Faithful GUI Agent Benchmarking

Tsinghua University

2

Submitted by

ben-dlwlrma

Representation over Routing: Overcoming Surrogate Hacking in Multi-Timescale PPO

·
1 authors

Submitted by

mazhdari

Broadening Access to Transportation Safety Data with Generative AI: A Schema-Grounded Framework for Spatial Natural Language Queries

umass

University of Massachusetts Amherst

Submitted by

alexicanesse

Decoupling Communication from Policy: Robust MARL under Bandwidth Constraints

orailix

Submitted by

CircleRadon

Pixel-Level Pavement Distress Assessment Using Instance Segmentation

·
5 authors

Submitted by

tianleliphoebe

ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison

·
10 authors

Submitted by

hoangp111

Decoding the Critique Mechanism in Large Reasoning Models

·
6 authors

Submitted by

shinying

MotiMotion: Motion-Controlled Video Generation with Visual Reasoning

adobe-research

Submitted by

YUKKKKKKKKKKKKK

Seeing the Needle in the Haystack: Towards Weakly-Supervised Log Instance Anomaly Localization via Counterfactual Perturbation

huggingface