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

Sep 21

Submitted by

zhangshaolei

EvoOntology: A Self-Evolving Ontology Layer for Data Agents

RUC-DataLab

Submitted by

tobiaslee

CodeMidas: Scaling Agentic Coding RL Environments from Code Itself

XiaomiMiMo

Submitted by

Yooki

Grounded Skill Synthesis from Code at Scale for Agentic Intelligence

ant-intl

ant-international

Submitted by

taesiri

RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents

Qwen

Submitted by

Xufew

IntBMoE: Integrating Block-Level Conditioning into Expert Composition for Full-Participation Mixture-of-Experts

·
5 authors

3

Submitted by

Harland

OmniVChat: Synthesizing, Benchmarking, and Training for Native Audio-Visual Dialogue

Qwen

Submitted by

sanaka87

Paint-Anything: Unified Any-Color Control for Image Generation and Editing

ByteDance-Seed

3

Submitted by

Hongyang-Du

Designer-RSI: Evolving Procedural Memory from User Traffic for Agentic Graphic Design

adobe

Submitted by

taesiri

OmniVBench: A Benchmark and Large-Scale Dataset for Omni Reference-to-Video Generation

tencent

Submitted by

chuxuan

BI-Agent and BI-Bench: Towards Automating End-to-End Business Intelligence

MicrosoftResearch

Microsoft Research

Submitted by

taesiri

MintAct: A Unified Visual Agent for Digital Environments

apple

Submitted by

zz1358m

GraphSkillEvo: Evolutionary Optimization of Graph-Structured Agent Skills

·
4 authors

Submitted by

nikhilsk

Gricea: An Open Science Platform for Conversational AI Research

JohnsHopkins

Johns Hopkins University

Submitted by

jojo23333

MoME: Mixture-of-Memory Embeddings for Context-Aware Sparse Lookup

vector-institute

Vector Institute

Submitted by

qqiang

Calibrating Teacher--Student Discrepancy for On-Policy Distillation

nanjinguniv

Nanjing University

Submitted by

Sichang0621

From Pretraining to Proficiency: Real-World Subtask RL for Long-Horizon Manipulation with Minimal Human Intervention

·
7 authors

2

Submitted by

Zihan1004

CADWorld: Computer-Use Benchmark for Long-Horizon Computer-Aided Design

·
7 authors

Submitted by

Q-M-E

MLLMs Hallucinate when Information Distribution Drifts in Synergy Heads

SUAT-SZ

Shenzhen University of Advanced Technology

Submitted by

taesiri

When AI Reviews Train AI Reviewers: Scientific-Judgment Collapse and Mitigation

·
3 authors

Submitted by

JimmyMa99

FRAUDSkill: Structured Frozen-Weight Skill Optimization for Audio Anti-Fraud Detection

PPSUCTeleantifraudCommunity

People’s Public Security University of China

Submitted by

JimmyMa99

TeleAntiFraud 2.0: A Refreshable, Profile-Grounded, and Audio-Based Benchmark for Telecom Fraud Detection

PPSUCTeleantifraudCommunity

People’s Public Security University of China

Submitted by

zhangzhongbo2213

Learning Foresight without Explicit Trajectories for 3D Diffusion Policies

dalian-university-of-technology

DaLian University of Technology

Submitted by

windchimeran

SiliconBench: Speed, Memory, and Fidelity for LLM Serving on Unified-Memory Desktops

Pennsylvania State University

Submitted by

Q-M-E

Training-Adaptive Convolutional Sparse Coding via Information Bottleneck for Robust Visual Representation

SUAT-SZ

Shenzhen University of Advanced Technology

Submitted by

Canlee

DeformSmith: Physics Harness-Guided Hierarchical Generation of Deformable Assets for Robot Manipulation

·
5 authors

Submitted by

utkarsh4430

SteerDuplex: Steerable Duplex Speech Dialogue Models

ScaleAI

Submitted by

ruiyangw

GAVEL: Graph World Models for Verified and Efficient Long-Horizon LLM Task Planning

DukeUniversity

Duke University

2

Submitted by

uchibeke

APort Vault: Benchmarking AI Agent Payment Authorization with the Open Agent Passport

aporthq

Submitted by

mura1n

Refinement Is Inherently Editable: Training-Free Prompt-to-Prompt Image Editing with Generative Refinement Network

CityU-HongKong

City University of Hong Kong

Submitted by

romerocruzsa

Retention-Constrained Post-Training Quantization of Cellpose-SAM for Stem Cell Microscopy

capicu-ai

3

Submitted by

utkarshagarwal

Geometry of Values: Task Vector Composition for Ethical Preference Alignment in Language Models

MBZUAI

Mohamed Bin Zayed University of Artificial Intelligence