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 9

Submitted by

chijw

AgentGarten: Code Worlds for Evolving Agents

MirroS-Lab

Submitted by

fuvty

TokenRouter: Efficient Serving System for Token-Level LLM Routing

nics-efc

Tsinghua-NICS-EFC

Submitted by

yyh929

SuperNav: An Agentic Navigation System for Any Task in Any Scene

Submitted by

whatseeker

MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement

XiaomiMiMo

Submitted by

enrue1893

Multi-Agent Egocentric World Model with Fine-Grained Embodied Interaction

kaist-ai

Submitted by

gothicwhw

In-context Robot Learning Made Simple: A Democratized Recipe for Manipulation Tasks

·
13 authors

Submitted by

Quanyu001

From Traces to Agentic Worlds: Agentic Language World Models for Interactive Environment Simulation

nanyang-technological-university-singapore

Nanyang Technological University Singapore

Submitted by

Abyssaledge

DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training

·
9 authors

Submitted by

wangzx1994

OuroWorld: Bringing Any 3D World Alive as Diverse, Endlessly Looping 3D Cinemagraphs

AlayaLab

Submitted by

zz1358m

Learn2Play Bench: How Well Do LLM Agents Learn from Experience in Unfamiliar Environments?

NationalUniversityofSingapore

National University of Singapore

Submitted by

henry-y1

MC-Sparse: Deconstructing and Closing the Dense-Sparse Attention Gap in Diffusion Transformers

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

CheeryLJH

TestPrism: Rethinking Test Evaluation Beyond a Single Reference

NJU-LINK

Submitted by

tbrx

U-Space: Uncovering When and Why Uncertainty Arises in Language Models

·
7 authors

Submitted by

banyh2000

Post-Training Frontier Text-to-Image Models by Composing Preference and Rubric Rewards

lmarena-ai

Submitted by

HaoyuZhao

Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks

UniversityCollegeLondon

University College London

Submitted by

Huan-WhoRegisteredMyName

SparseDecoding: Decoding-Aware Pruning for Accurate and Efficient LLM Inference

encodelab

Westlake ENCODE Lab

1

Submitted by

suhwan-cho

LEGO: A Lifting-Free Approach for Exocentric-to-Egocentric Video Generation

25frms

Submitted by

zpy777

Reasoning-Informed Visual Editing

VisionXLab

SJTU VisionXLab

Submitted by

stceum

Foundations of Large Language Models

NiuTrans

Submitted by

appletea2333

OneSearch-VL: Unified Multimodal Deep Research Agent for Image and Video

·
14 authors

Submitted by

Jeryi

What Did the Agent Actually Do? Evidence-Grounded Oversight for Long-Horizon Agents

RUC-GSAI-IIRLab

RUC-GSAI-IIRLab

Submitted by

Jinjing713

VibeEdit: Image Editing with Canvas Instructions

Sydney-Uni

The University of Sydney

Submitted by

JitaiHao

SparseEngine: Sparse-First Inference Engine

·
4 authors

Submitted by

xingyoujun

USDCraft: Geometrically Grounded Programmatic Modeling of Articulated 3D Assets for Simulation

nvidia

1

Submitted by

yzzyu

OmniCapBench: A Deep-Structured Evaluation Framework for Fine-Grained Audio-Visual Captioning

Tencent-Hunyuan

Tencent Hunyuan

1

Submitted by

KairuiHu

Embodied Turing Machines: Stateful Code for Robot Recursive Self-Improvement

·
5 authors

Submitted by

hongxingli

ViSkill: Reinforcing VLM Agents with Evolving Visual-Native Skills

zju

Zhejiang University

Submitted by

Shuyu12138

SanSi: A Looped Typed Decision Model for System 1.5 Thinking

minnesotanlp

Submitted by

Lutalica

V-CoLA: Vision Token Compression with Linear Attention

ATH-MaaS

1

Submitted by

YhangChen

ReSPO: Reshaped Sequence Policy Optimization for Gradient Starvation in Off-Policy Learning

UCLA Computational Machine Learning Lab

Submitted by

teddybearagee

Do LLMs Understand Sequential Structure? A Controlled Study of Inference and Generation

illinois

University of Illinois at Urbana-Champaign

1

Submitted by

SnowflakeWang

From Prompting to Composing: A Spatial Canvas Interface for Poster Generation

microsoft

Submitted by

hongxingli

SpaceCast-Bench: Evaluating Predictive Spatial Reasoning in Vision-Language Models

zju

Zhejiang University

Submitted by

bowiehsu

SpatialOPSD: Self-Distilling Spatial Intelligence from Verified Coding Agent Traces

ATH-MaaS

1

Submitted by

schlegelm

Retrieval-Centric Deep Learning in Growing Nonparametric Neural Networks

google

Submitted by

ankanmbz

WorldGuide: Goal-Directed Video World Model for Procedural Task Execution

MBZUAI

Mohamed Bin Zayed University of Artificial Intelligence

Submitted by

luping-liu

Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching

hkuhk

The University of Hong Kong

Submitted by

hmhm1229

Chaos in the Text: Revealing the Modality Preference in Mixed-Modality Retrievers

·
9 authors

Submitted by

neil-ai

SpaceFlow: Locally Controllable 3D Generation

ethz

Submitted by

mocoV3

REMORY: Learning Residual Memory for Context Compaction

·
8 authors

Submitted by

KaiserWhoLearns

Accurate but Not Humble: Evaluating Epistemic Humility in LLM Agents under Knowledge Conflict

JohnsHopkins

Johns Hopkins University

Submitted by

pengxiang

Scaling to Tens of Thousands of Test-Time Iterations with Loop-Native Attention Residuals

·
6 authors

Submitted by

adrianb1

One Block, Multiple Depths: Recurrent Vision Transformers with Depth-Programmed Experts

SamsungResearch

Samsung Research

Submitted by

trxcc2002

A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization

·
11 authors

Submitted by

ashutosh1919

Synthesis Through Simulation: Generating Coherent Enterprise Data via Scalable Agent-System Interaction

SAP

1

Submitted by

stephenchungmh

Can AI Agents Make Open-Ended Scientific Discovery? Evidence from Station

TheHKU

Hong Kong University

Submitted by

Rain729

Pumpire: Unified Benchmark for Metric Distance Estimation

·
9 authors

Submitted by

mario-sanz

Investigating the Role of Reasoning-Language Alignment in Monolingual Retrieval-Augmented Generation

·
3 authors

1

Submitted by

Franck-Dernoncourt

EDiS: Edge Disjoint Subgraph Sparsification Framework for Graph Neural Networks

·
10 authors

Submitted by

lr10260

CARE: Certifying Acceleration for Vision-Language-Action Inference

UMCP

University of Maryland College Park

Submitted by

yixuantt

On-Policy Distillation Teaches New Skills but Not New Knowledge

·
2 authors

Submitted by

kulits

BrickBench: Evaluating Agentic Brick Design

·
5 authors

Submitted by

hongxingli

Distilling Routed 3D Privilege for Spatial Reasoning in Vision-Language Models

zju

Zhejiang University

Submitted by

chungen

SpecFold: Folding Multi-Branch Redundancy for Faster Speculative Decoding in Diffusion Language Models

GeorgiaTech

Georgia Institute of Technology

Submitted by

KKKKAAAAIIII

MIRA: A Musical Intent Refinement Agent for Aligning Text-to-Music Generation with User Intent

·
5 authors

Submitted by

merlinarer

Incremental Open-Ended Deep Research with Structured Harness

·
2 authors

2

Submitted by

Junle-cs

You Changed Your Mind, The Model Didn't: Demystifying Intent in Multi-Turn Dialogue

·
8 authors

Submitted by

caiyuchen

Learning to Steer, Steering to See: Unveiling the Geometry of RLVR in Large Language Models via Trainable Vectors

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

Eku127

SatNav: A Scalable Benchmark for Long-Horizon UAV Vision-Language Navigation from Satellite Imagery

·
7 authors

Submitted by

jarmstrong001

MARGIN: Runtime Confidence Calibration for Multi-Agent Foundation Model Coordination

EricssonAB

1

Submitted by

GhostCai

SPW-Nav: A Streaming Panoramic World Model for Language-Guided Navigation

AlayaLab

Submitted by

ShyFoo

TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows

the-aiml

Australian Institute for Machine Learning