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

May 21

Submitted by

xwm

Video2GUI: Synthesizing Large-Scale Interaction Trajectories for Generalized GUI Agent Pretraining

PekingUniversity

Peking University

Submitted by

filicos

Mega-ASR: Towards In-the-wild^2 Speech Recognition via Scaling up Real-world Acoustic Simulation

NationalUniversityofSingapore

National University of Singapore

Submitted by

xiaochonglinghu

Enhancing Train-Free Infinite-Frame Generation for Consistent Long Videos

alibaba-inc

Submitted by

weizhepei

You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectories

·
6 authors

Submitted by

ZhenlongYuan

IndusAgent: Reinforcing Open-Vocabulary Industrial Anomaly Detection with Agentic Tools

ict-cas

Institute of Computing Technology, Chinese Academy of Sciences

1

Submitted by

zunhai

OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond

hkuhk

The University of Hong Kong

Submitted by

AustinXiao

A Survey of Large Audio Language Models: Generalization, Trustworthiness, and Outlook

NanyangTechnologicalUniversity

Nanyang Technological University

Submitted by

wgcyeo

It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs

kaist-ai

Submitted by

Emaad

Toto 2.0: Time Series Forecasting Enters the Scaling Era

Datadog

Submitted by

haiquanlu

Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs

NationalUniversityofSingapore

National University of Singapore

Submitted by

jojo0217

Generative Recursive Reasoning

·
6 authors

1

Submitted by

HelenMao

CutVerse: A Compositional GUI Agents Benchmark for Media Post-Production Editing

CUC-MIPG

Multimedia Intelligent Processing Group in Communication University of China

Submitted by

taesiri

Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning

·
7 authors

Submitted by

konglongge

LLMEval-Logic: A Solver-Verified Chinese Benchmark for Logical Reasoning of LLMs with Adversarial Hardening

llmeval-fdu

LLMEval Official Team

Submitted by

imone

HRM-Text: Efficient Pretraining Beyond Scaling

sapientinc

Submitted by

seungone

On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientists

CarnegieMellonU

Carnegie Mellon University

Submitted by

DhavalPatel

Evaluating Temporal Semantic Caching and Workflow Optimization in Agentic Plan-Execute Pipelines

ibm

1

Submitted by

yli-ml

Learning from Language Feedback via Variational Policy Distillation

Salesforce

Salesforce AI Research

Submitted by

mboss

OCTOPUS: Optimized KV Cache for Transformers via Octahedral Parametrization Under optimal Squared error quantization

stabilityai

1

Submitted by

HenghuiDing

OcclusionFormer: Arranging Z-Order for Layout-Grounded Image Generation

FudanCVL

Submitted by

gohyojun15

Stitched Value Model for Diffusion Alignment

·
11 authors

Submitted by

nielsr

Stable Audio 3

stabilityai

Submitted by

Mehrab-Tanjim

MOCHA: Multi-Objective Chebyshev Annealing for Agent Skill Optimization

adobe-research

Submitted by

le723z

RiT: Vanilla Diffusion Transformers Suffice in Representation Space

mila-intel

Submitted by

Haotian-sx

UniT: Unified Geometry Learning with Group Autoregressive Transformer

Submitted by

YuZeng260

SaaSBench: Exploring the Boundaries of Coding Agents in Long-Horizon Enterprise SaaS Engineering

·
14 authors

Submitted by

joykirat

MINTEval: Evaluating Memory under Multi-Target Interference in Long-Horizon Agent Systems

·
6 authors

1

Submitted by

JiaJinrang

PanoWorld: A Generative Spatial World Model for Consistent Whole-House Panorama Synthesis

·
4 authors

1

Submitted by

cyl

The Unlearnability Phenomenon in RLVR for Language Models

·
3 authors

Submitted by

visity

Conditional Equivalence of DPO and RLHF: Implicit Assumption, Failure Modes, and Provable Alignment

·
6 authors

Submitted by

taesiri

PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

taesiri

SpecBench: Measuring Reward Hacking in Long-Horizon Coding Agents

·
4 authors

Submitted by

taesiri

Mem-π: Adaptive Memory through Learning When and What to Generate

·
8 authors

Submitted by

aradianis

Learn-by-Wire Training Control Governance: Bounded Autonomous Training Under Stress for Stability and Efficiency

QluonAI

Submitted by

taesiri

DrawMotion: Generating 3D Human Motions by Freehand Drawing

·
10 authors

Submitted by

wark123

Capturing LLM Capabilities via Evidence-Calibrated Query Clustering

Uwmadison

University of Wisconsin-Madison

1

Submitted by

gigant

Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation

NousResearch

Submitted by

long2333

Safety Alignment as Continual Learning: Mitigating the Alignment Tax via Orthogonal Gradient Projection

Tsinghua University

Submitted by

zxbsmk

iTryOn: Mastering Interactive Video Virtual Try-On with Spatial-Semantic Guidance

·
9 authors

1

Submitted by

gzxiong

Rethinking Visual Attribution for Chest X-ray Reasoning in Large Vision Language Models

·
5 authors

Submitted by

imarod

TIDE: Efficient and Lossless MoE Diffusion LLM Inference with I/O-aware Expert Offload

·
5 authors

Submitted by

Kirscher

Lost in the Folds: When Cross-Validation Is Not a Deep Ensemble for Uncertainty Estimation

MIC-DKFZ

Submitted by

wark123

DynMuon: A Dynamic Spectral Shaping View of Muon

Uwmadison

University of Wisconsin-Madison