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

Jul 27

Submitted by

lhpku20010120

DataPrep-Bench: Benchmarking LLMs as Training Data Preparators

PekingUniversity

Peking University

Submitted by

taesiri

Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills

QwenBusinessUnit-Edu

QwenBusinessUnit-Edu

Submitted by

taesiri

Molt: A Scalable PyTorch-Native Training Framework for Agentic Reinforcement Learning

nvidia

Submitted by

gdadhich

Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems

MaximemAI

Submitted by

taesiri

Scaling Native Multimodal Pre-Training From Scratch

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

yuntian-deng

Interactive Training 2: Auditable Control Plane for Live Model Training

UWaterloo

University of Waterloo

Submitted by

sp12138sp

Three-Body Scattering for Generative Modeling

·
6 authors

Submitted by

hongst

LAMAR: An Open Language-Aware Multilingual Alignment Reranker

nlpai-lab

NLP & AI - Korea University

Submitted by

yuanmei424

O-VAD: Industrial Video Anomaly Detection through Object-Centric Tracking and Reasoning

CarnegieMellonU

Carnegie Mellon University

Submitted by

AmirhoseinGH

Multi-Head Latent Control: A Unified Interface for LLM Agent Decision Making

huawei technologies co. ltd

Submitted by

taesiri

Closing the Loop: Training-Free Revisit Consistency for Autoregressive Generative Rendering

·
4 authors

Submitted by

soujanyaporia

IDEAgent: Agentic Quality-Diversity Search for Research Idea Generation

declare-lab

Deep Cognition and Language Research (DeCLaRe) Lab

Submitted by

taesiri

SceneActBench: Can Agents Act on the 3D Scenes They See?

·
14 authors

Submitted by

u-kob

Spectral Prior for Reducing Exposure Bias in Diffusion Models

Sony

Submitted by

YuanchengXu

ID-V2V: Identity-Preserving Video Restylization

netflix

Submitted by

zpatrick

VisCo: Leveraging Large Language Models as Intrinsic Encoders for Visual Token Compression

·
4 authors

Submitted by

ash56

Multimodal Speaker Verification as a Threat to Speaker Anonymization

JohnsHopkins

Johns Hopkins University