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

Nov 4

Submitted by

IntJudge

Route Anything: A Study of Model and Policy Routing for General-Purpose AI Systems

·
7 authors

Submitted by

ElsaShaw

Words Caught in the Heart: Simulating Person-Specific Expression for Self-Opaque Clients in Psychotherapy

NextGenWhu

Submitted by

lxucs

Think Less, Locally: Local Preference Learning for Reasoning Compression

tencent

Submitted by

hyeonho-jeong-video

MoviEdit: Long-form Multi-Shot Video Editing Made Easy

adobe

Submitted by

LongoXC

Behind Video Generation: Observations and Insights on Multimodal Understanding from Kling

KlingTeam

KlingAI Research

Submitted by

chyun

DreamX: Video Diffusion Model is a Universal Latent Action Learner

kaist-ai

Submitted by

Jiafei1224

RoboColosseum: Evaluating Physical Intelligence at Scale In-The-Wild

huggingface

Submitted by

WJ0830

Factorized MLLM Embeddings for Composed Image Retrieval

kaist-ai

Submitted by

ChrisBullet

Axion 1.0: Open Agent Foundation Models, Simplified

·
7 authors

Submitted by

dongguanting

ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents

RUC

Renmin University of China

Submitted by

HarrisonPENG

YANAMI: A Large-Scale Multimodal Dataset and Benchmark for Multiturn Stateful Interaction

·
13 authors

1

Submitted by

Lruca

VidUEU-Agent: A Data-Curation Agent for Multimodal Understanding, Editing, and Unified Tasks

·
15 authors

Submitted by

Qing145

EvoTeam: Multi-Agent Collaboration via Evolving Teams

JLU-Deep

DEEP Group at Jilin University

Submitted by

ChilleD

WebHarbor: Docking Real Websites for Evolving GUI Agent Environments

microsoft

Submitted by

yikuuun

MoLRSI: A Recursive Self-Improving AI agent for Molecular Discovery

·
18 authors

1

Submitted by

stargazerx0

SpecPilot: Autonomous Algorithm Discovery and Systems Optimization for Speculative Decoding

UCBerkeley

University of California, Berkeley

Submitted by

JueZhang

ParetoSaddler: Pareto-Guided Optimization of Agent Harnesses for Quality and Cost

microsoft

Submitted by

WWeiheng

"Shall We Approve?" Consequence-Aware Approval for Autonomous Coding Agents

SingaporeManagementUniversity

Singapore Management University

Submitted by

xyzhang626

Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use From Evolved Harnesses

microsoft

Submitted by

niceirene

Recursa: Toward Recursively Self-Improving Foundation Model Systems

alibabagroup

1

Submitted by

Yesianrohn

FlowTranslator: Training-Free In-Image Machine Translation with Flow-Based Editing

Fudan-University

Fudan University

1

Submitted by

pbelcak

You don't need long context

nvidia

Submitted by

travisddavies

PrimeSim: A Generalist World Model Simulator from Large-Scale Cross-Embodiment Demonstrations

·
13 authors

Submitted by

happyiai

Every Token Needs Its Own Teacher: Multi-teacher Adaptive On-policy Self-distillation

TUM

Technical University of Munich

Submitted by

BBBBCHAN

Near-Zero Token Embeddings Silently Stall Training in Gated Sequence Models

·
6 authors

1

Submitted by

luohy

Adversarial Speculative Reasoning Improves Long Horizon Agents

SubconsciousDev

1

Submitted by

shantanuacharya

Dory1.0-4B: A Looped Normalized-GPT

nvidia

1

Submitted by

Mqleet

Switch-SWE: Do Coding Agents Hold Focus When Switching Between Unfinished Tasks?

SJTU

Shanghai Jiao Tong University

Submitted by

MisakiWang

Can AI Reviewers See Your Efforts to Improve Your Paper? A Large-Scale Study of Paper Revision Trajectories

umd-zhou-lab

Submitted by

zjy2001

LOLPlay: Benchmarking Real-Time Visual Agents in League of Legends

·
1 authors

Submitted by

RyanL22

RoboTryOn: Synthesizing Dexterous Robot Demonstrations from Human Videos via Virtual Try-On

kaist-ai

Submitted by

Jethro37

Are Frontier VLM Agents Ready to Be Robot Generalists? An Empirical Study with the Embodied Agent Arena

·
10 authors

Submitted by

XiaokunSun

SeMo3D: Unifying 3D Generation, Editing, and Completion via Selective Modeling

nanjinguniv

Nanjing University

Submitted by

yhsure

BEND 2.0: Benchmarking the Capabilities Needed to Design DNA

·
9 authors

Submitted by

marcoshernanz

Muon Makes the First Mixture-of-Experts Layer Harder to Balance

·
1 authors

Submitted by

moein99

Every Pixel Is an Option: A System One Model for Medical Ultrasound Biometry

UBC-V

University of British Columbia

Submitted by

Parv17k

Do We Need an LLM to Know Why Sources Conflict? Decision Models for Knowledge-Conflict Typing in RAG

·
4 authors

Submitted by

AlanRosston520

Is Your Long-Horizon Benchmark Really Deep? Measuring and Synthesizing Depth in Terminal Tasks

·
9 authors

Submitted by

drezzaali

What We Keep of What We Forget: Replacement Operators and Routing Semantics in Training-Free MoE Expert Pruning

·
4 authors

1

Submitted by

SP4595

Is Forgetting Unified? Exploring Concept Unlearning in Unified Multimodal Models

CASIA

Chinese Academic of Science Institute of Automation

1

Submitted by

Yuxiang007

Where to Learn and How Far to Roll Out: Efficient Adaptive Training for GUI Agents

·
5 authors

Submitted by

Jinnkunn

Cross-Prompt Reuse of Entity-Replacement Responses

·
4 authors

Submitted by

WTHH031230

SphereFlow: Missing Modality Imputation via Geometric Transport on Hypersphere

·
9 authors

Submitted by

ovcompcase

Kierston & KierstonBench: Simulation of Allentown, PA and Evaluation of Language Models in Crime Investigation and Surveillance Tasks

·
2 authors

Submitted by

yfyeung

GigaSpeech 3: Scaling Multi-Dialect Arabic Speech Recognition with Weakly Labeled Data

X-LANCE

SJTU Cross Media Language Intelligence Lab

Submitted by

ZZL2005

Auto-Research Agent on MSA Organization to Improve Protein-Protein Complex Structure Prediction

SJTU

Shanghai Jiao Tong University

1

Submitted by

akhadangi

Going Deeper With Memories

·
1 authors

Submitted by

suusuu93

Enhancing Stock Market Information Retrieval with Retrieval-Augmented Generation

·
6 authors

Submitted by

leizhao7

When Fine-Grained Credit Assignment Helps and Hurts: Multimodal Agentic Reinforcement Learning through Video Editing

philo-labs

Submitted by

yliusci

Length Is Cheap, Competition Is Not: Attention Predicts Context Rot in Retrieval

·
1 authors

Submitted by

elixpo

QFabric: Guarded Cross-Domain Task Recovery and Its Limits on a Linux-RTOS Board

elixpoo