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 30

Submitted by

SempraETY

Intrinsic Distribution Matching

inclusionAI

Submitted by

zhshj0110

A-RSI: Adaptive Task Refresh for Zero-Data Multimodal Self-Improvement

xiaomi-research

Xiaomi Research

3

Submitted by

keyangx3

Your Benchmark Is an Empirical Measure Over Difficulty

UIUC-CS

University of Illinois at Urbana-Champaign

Submitted by

hyeon9698

Look Only When Needed: Certificate-Guided Visual Routing for Efficient Diffusion VLMs

kaist-ai

Submitted by

wlsgur1018

L3G-World: Real-Time Infinite World Generation with Latent 3D Gaussian Memory

kaist-ai

Submitted by

Tingyang-Chen

WorkFactory: Turning Professional Intents into Executable Work for Scalable Agent Training

IQuestLab

Submitted by

MindscapeRAG

A Broader Look into MLLM's Retrieval Heads

tencent

Submitted by

Gokottaw434

Read, Allocate, Route: Role-Coupled Tokenization for Pixel-Space Generative Transformers

·
5 authors

Submitted by

shenyunhang

Youtu-Parsing-Omni: One Encoder, One Schema, All Modalities

tencent

Submitted by

Wiselnn

Co-Design Bench: Multimodal Collaborative Coding Evaluation for Agentic Models

Qwen

Submitted by

void-2024

VibeGame: Prompt-to-Game Development with AI-Native Engine and Self-Evolving Adversarial Agent Team

nanjinguniv

Nanjing University

Submitted by

hyz317

HairDeck: Animation-Ready Stylized Hair Cards from a Single Image

Tsinghua University

Submitted by

Yesianrohn

WeVisCut: The Whole Editing Room in One Agent From Narration to Long-Form Video

tencent

Submitted by

yushaohan

IntReal: A Real-World Dataset and Benchmark for Multimodal Interleaved Generation

meituan-longcat

Submitted by

vyokky

WebAnything: From the Wild Web to Scalable Agent Environments

·
7 authors

Submitted by

D-YZ

CoSTA: Compressed Speech Token Adaptation for Efficient Spoken Dialogue Models

CASIA

Chinese Academic of Science Institute of Automation

Submitted by

metinovadilet

Glance: The Generator Is Its Own Reranker

airun-ai

Airun Intelligence Labs

1

Submitted by

caiyuchen

Investor's Last Exam: Can AI Agents Do Equity Investment Without Seeing the Future?

·
10 authors

Submitted by

wanglucs

AskStream: When and What to Ask Proactively under Uncertainty in Streaming Video Understanding

CASIA

Chinese Academic of Science Institute of Automation

Submitted by

xuhaiyang3110

EditAgent: Rethinking Agentic Image Editing

adobe

Submitted by

appletea2333

PhysInsert: Physically Plausible Object Insertion with Geometric Priors and Bidirectional Object-Scene Interactions

·
8 authors

Submitted by

JaydenTeoh

Multi-Token Prediction Done Right

Massachusetts-Institute-of-Technology

Massachusetts Institute of Technology

Submitted by

RandallBalestriero

Proving and Improving the Platonic Representation Hypothesis with JEPAs

galilai-group

Submitted by

Ideallll

OpenFlyV2: An onboard vision-language-action model for aerial perception and navigation

IPEC-COMMUNITY

IPEC at Shanghai AI Laboratory

Submitted by

xubinknight

AB-Attention: Accurate Block Attention for Accelerating Video Diffusion Transformers

·
9 authors

Submitted by

peitongd

Online Dense Reinforcement Learning with Planning-Grounding Agent for UI Navigation on Live IOS Apps

apple

Submitted by

CauchyLovesU

Your LR Schedule May Not Be Principled: Effective Learning Rate Scheduling in Weight-Normalized Training

ByteDance-Seed

Submitted by

5hadytru

SO-101 Bench: Accessible Spatial Manipulation

·
1 authors

Submitted by

xhyumiracle

ABF-Bench: A Verifiable Benchmark for Agentic Blockchain Forensics

ImperialCollegeLondon

Imperial College London

Submitted by

yslan

Holo4D: Holistic 4D Reconstruction as Geometric Control for Video Diffusion

Oxford

University of Oxford

Submitted by

ZouBo

After One-Click Video: The Inflow of Service-Type Content, the Failure of the Recommendation Pool Mechanism, and the BoLe System

BreadStudio

1

Submitted by

AZH04

Training Foundation Models with a Single Algorithm

CarnegieMellonU

Carnegie Mellon University

Submitted by

Debargha

Scaling Lean Theorem Proving for Real-World Code Verification

amazonwebservices

Amazon Web Services (AWS)

Submitted by

ShineiArakawa

FDSML: Formula-Driven Supervised Metric Learning with Parameter-Aware Triplet Loss

·
6 authors