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

Oct 19

Submitted by

caizhongang

Has GPT-6-Astra Achieved Embodied Intelligence?

sensenova

Submitted by

callmeethan

Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents

McAuley-Lab

Submitted by

HongchengGao

HexaShell: A Thin Harness for Exploring the Limits of VLM Agents as Robot Coding Agents

HexaFuture

Submitted by

seungone

Into the Unknown: Benchmarking AI agents on open scientific research problems

CarnegieMellonU

Carnegie Mellon University

Submitted by

Yooki

BizBench: Unveiling Hidden Challenges in Deploying Long-Horizon Agents for Real-World Business

ant-intl

ant-international

Submitted by

williamium

Learning to Think and Explore in 3D World

UCSanDiego

University of California at San Diego

Submitted by

f66

BridgeWAM: Bridging Video and Action Experts via Latent Bridge Queries

Fudan-University

Fudan University

Submitted by

cckevinn

OpenMobile-2: Building Versatile Mobile Agents with Scalable Environments and App-Native Tools

NJU

Nanjing University

Submitted by

ramiroluo

From Research Goals to Scientific Closure: Evaluating Agents Through Evidence-Supported Physics Research

Shanghai-AI-Laboratory

Shanghai AI Laboratory

Submitted by

gfyddha

Tree-Aware Posterior Distillation for Long-Horizon Agentic Reinforcement Learning

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

hyb1124

Speculative Sampling in Interactive Video Generation

ustc

university of science and technology of china

Submitted by

MathsCode

DLite: History-free Speculative Decoding for Agentic AI

SJTU

Shanghai Jiao Tong University

Submitted by

VincentHancoder

RAP: A Dataset for Paper-to-Poster Generation via Declarative Markup from Human-Crafted Exemplars

meituan

Submitted by

Wayne-King

Execution Before Memory: Diagnosing Memory Failures in Video World Models

hkuhk

The University of Hong Kong

Submitted by

Lingaaaaaaa

DiscoveryRSI: Recursively Learning to Discover Valuable Hypotheses

PhAILabs

Submitted by

hcwei

OfficeTown: Generative Organizations for Evolving Agent Intelligence

microsoft

Submitted by

jianguda

Lexical Reference Frames: Interpreting LM Residual Streams in Native Semantics

·
4 authors

Submitted by

sun1245

SciΠvot: Benchmarking Principal-Investigator Judgment at Pivotal Moments of Scientific AutoResearch Agents

·
18 authors

Submitted by

Dvir

Stuck on Shuffle: On Attribute-Object Binding in Vision-Language Models

·
7 authors

Submitted by

lwk21

Can Visual Generation Expand What Unified Multimodal Models Know?

zai-org

Submitted by

Phospheneser

MOSS-Image Technical Report

OpenMOSS-Team

Submitted by

YYangzzzz

OS-Helix: Unifying GUI Interaction and Code Execution in Native Computer-Using Agents

ustc

university of science and technology of china

Submitted by

Tsuki0512

AVC-Reward: Event-Centric Audio-Video Consistency Evaluation

Zhejiang University

Submitted by

ziyan2003

MorphAct: Robot Policies Morphing with the Worlds They Act In

SJTU

Shanghai Jiao Tong University

Submitted by

f20180301

DuoDiscover: Scaling Autonomous Discovery Beyond the Evaluation Bottleneck

google

Submitted by

ZBox008003

Dream2Discern: Learning from Hard Latent Variants for Runtime Failure Prediction in Multi-Agent Systems

RutgersU

Rutgers University

Submitted by

QLNI

Embeddings Are Overpriced? A Frozen 16 MB Vocabulary for Small and Ternary Language Models

·
1 authors

Submitted by

jonathanShi

OPMD: On-Policy Mechanism Distillation Beyond Output Supervision

ShanghaiTech

ShanghaiTech University

Submitted by

songdj

PremiseCheck: Efficient Fact-Checking of Human and AI Peer Reviews

·
5 authors

Submitted by

HuskyDoge

Towards Looped Models Done Right, Part I: Topology, Input Injection, Recurrent-State Design

IFM

Institute of Foundation Models

Submitted by

liuqa

Game2Real-Bench: Do Toy Games Reveal and Transfer to Real-World Visual Reasoning?

·
9 authors

Submitted by

wuxiyang

Failure Recovery via Belief Updates and World Modeling for Long-Horizon Agents

IntelligenceLab

Agent Intelligence-Lab 🦞

Submitted by

fujial

FC-SWE: Failure-Conditioned RL for Long-Horizon Software Engineering Agents

nvidia

2

Submitted by

jhying

Belief-Trajectory Energy: Measuring the Path to a Prediction

·
10 authors

1

Submitted by

xiaocong01

Skip the Compressor: Coding Agents Fix More Issues When They Pick Whole Files

AI-Interpretability-illinois

AI Interpretability @ Illinois

Submitted by

songdj

LEV: Thinking Fast and Deep with Looped Decision Models

·
4 authors

Submitted by

xiaocong01

Resisting Harmful Instructions through Training-Free Weight Editing

AI-Interpretability-illinois

AI Interpretability @ Illinois

Submitted by

xiaocong01

DAGFormer: Global Topology Prediction for Transformer Cross-Layer Connections

AI-Interpretability-illinois

AI Interpretability @ Illinois

Submitted by

haoyi-yang

Selection Is Not Learning: Choosing Model Futures

MBZUAI

Mohamed Bin Zayed University of Artificial Intelligence

1

Submitted by

zunhai

ReaDy: Recurrent States as Dynamic Context for Speculative Drafting in Hybrid Linear Attention Large Language Models

·
6 authors

Submitted by

hyz317

Persistent Planar Memory for Video World Models

Robbyant-Research

Robbyant Research

Submitted by

mhroh01

Dusting the Phantom: Gaussian Allocation for Fading Dynamic Content

·
7 authors

1

Submitted by

xiaocong01

Sparse Autoencoders Are a Shared Coordinate System for Symbolic Matching

AI-Interpretability-illinois

AI Interpretability @ Illinois

Submitted by

Wayne42

DuET: Note-Frame Dual Encoding for Piano Performance-to-Score Transcription

NationalUniversityofSingapore

National University of Singapore

Submitted by

Gaaaavin

StopWatch: Goal Grounding for Object-goal Navigation

Amazon-FAR

Amazon Frontier AI & Robotics

Submitted by

tuandattttt

PRISM: PRIvileged Sibling-guided Multi-Agent Topology Design

NationalUniversityofSingapore

National University of Singapore

Submitted by

fyxfelixfu

Consistent Distribution Matching for Data-Free Diffusion Distillation

vector-institute

Vector Institute

Submitted by

HongchangLiu

GeoState: Learning View-Decoupled Geometry States for 3D World Modeling

Tongji Unversity