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

Jun 23

Submitted by

JiayuJeff

PlanBench-XL: Evaluating Long-Horizon Planning of LLM Tool-Use Agents in Large-Scale Tool Ecosystems

UIUC-CS

University of Illinois at Urbana-Champaign

Submitted by

smallkang2025

OpenRath: Session-Centered Runtime State for Agent Systems

·
3 authors

Submitted by

Coneonewan

DataClaw0: Agentic Tailoring Multimodal Data from Raw Streams

·
9 authors

Submitted by

iseesaw

EnterpriseClawBench: Benchmarking Agents from Real Workplace Sessions

FrontisAI

Submitted by

vishesh-t27

Grouped Query Experts: Mixture-of-Experts on GQA Self-Attention

FrontiersMind

Submitted by

Yuki131

KaLM-Reranker-V1: Fast but Not Late Interaction for Compressed Document Reranking

KaLM-Embedding

2

Submitted by

florinshum

World Action Models: A Survey

NationalUniversityofSingapore

National University of Singapore

Submitted by

CheeryLJH

CLI-Universe: Towards Verifiable Task Synthesis Engine for Terminal Agents

NJU-LINK

Submitted by

BradyFU

EvoEmbedding: Evolvable Representations for Long-Context Retrieval and Agentic Memory

nanjinguniv

Nanjing University

Submitted by

QizhiPei

BioMatrix: Towards a Comprehensive Biological Foundation Model Spanning the Modality Matrix of Sequences, Structures, and Language

Submitted by

tzt

HydraHead: From Head-Level Functional Heterogeneity to Specialized Attention Hybridization

·
7 authors

1

Submitted by

XUANMINGZHANG

Deeper is Not Always Better: Mitigating the Alignment Tax via Confident Layer Decoding

Qwen

Submitted by

yurun-chen

SkillHarness: Harnessing Safe Skills for Computer-Use Agents

Zhejiang University

Submitted by

shenyao

Learning from Your Own Mistakes: Constructing Learnable Micro-Reflective Trajectories for Self-Distillation

QwenBusinessUnit-Edu

QwenBusinessUnit-Edu

1

Submitted by

taesiri

Unlimited OCR Works

baidu

Submitted by

oldTOM

Foresight: Failure Detection for Long-Horizon Robotic Manipulation with Action-Conditioned World Model Latents

umich

University of Michigan

Submitted by

yigengx

Deep Research in Physical Sciences: A Multi-Agent Framework and Comprehensive Benchmark

Submitted by

taesiri

Self-Compacting Language Model Agents

·
8 authors

Submitted by

taesiri

Training Open Models for Agentic Phone Use

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

linyiyang2023

Notes2Skills: From Lab Notebooks to Certainty-Aware Scientific Agent Skills

SUSTech

Southern university of science and technology

1

Submitted by

taesiri

Tmax: A simple recipe for terminal agents

·
6 authors

Submitted by

wagoriginal

DailyReport: An Open-ended Benchmark for Evaluating Search Agents on Daily Search Tasks

·
11 authors

Submitted by

cheese1

Exploring the Design Space of Reward Backpropagation for Flow Matching

tencent

Submitted by

jaehong31

Safe Few-Step Generation via Velocity Editing

nanyang-technological-university-singapore

Nanyang Technological University Singapore

Submitted by

arkilpatel

Toward Open Weight Models Without Risks: Separating Public and Private Capabilities in LLMs

MilaQuebec

Mila – Quebec Artificial Intelligence Institute

Submitted by

quiet-storm

PoLAR: Factorizing Extent and Mode in Latent Actions for Robot Policy Learning

SeoulNatlUniv

Seoul National University

Submitted by

yanxi-chen

Connect the Dots: Training LLMs for Long-Lifecycle Agents with Cross-Domain Generalization Via Reinforcement Learning

AlibabaTongyiLab

1

Submitted by

yujiazheng

Causal Discovery in the Era of Agents

CarnegieMellonU

Carnegie Mellon University

Submitted by

taesiri

Tapered Language Models

·
3 authors

Submitted by

spisupat

Counsel: A Meta-Evaluation Dataset for Agentic Tasks

AtlaAI

1

Submitted by

yzhuoning

Vera: A Layered Diffusion Model for Content-Preserving Video Editing

netflix

1

Submitted by

jihochoi

Dense Reward for Multi-View 3D Reasoning with Global Maps and Local Views

kaist-ai

Submitted by

chenfeng1271

PolicyTrim: Boosting Intrinsic Policy Efficiency of Vision-Language-Action Models

sichuan university

Submitted by

harshpatel2898

A Verifiable Search Is Not a Learnable Chain-of-Thought

·
1 authors

Submitted by

darrienmckenzie

Manifold Bandits: Bayesian Curriculum Learning over the Latent Geometry of Large Language Models

UCSanDiego

University of California at San Diego

Submitted by

zfying

Comparing Linear Probes with Mahalanobis Cosine Similarity

·
3 authors

1

Submitted by

thrshr

FastMix: Fast Data Mixture Optimization via Gradient Descent

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

JDihlmann

Arbor: Explicit Geometric Conditioning for Controllable 3D Asset Generation

StabilityLabs

Submitted by

qsun2001

MeshFlow: Mesh Generation with Equivariant Flow Matching

·
9 authors

Submitted by

g-astruc

UniverSat: Resolution- and Modality-Agnostic Transformers for Earth Observation

·
5 authors

Submitted by

thebluser

Lift4D: Harmonizing Single-View 3D Estimation for 4D Reconstruction In-the-Wild

·
7 authors

Submitted by

zhenwang9102

Demystifying Training-Time Augmentation for Data-Constrained Language Model Pretraining

UCSanDiego

University of California at San Diego

Submitted by

ashwinnv

CalVerT: Augmenting Agents with Calibrated Verifier Telemetry Improves Action and Learning in Knowledge-Intensive Tasks

UTEXAS

University of Texas at Austin

Submitted by

kmcode

Go-with-the-Track: Video Compositing and Motion Control with Point Tracking

Eyeline-Labs

Submitted by

MatanBT

TROPT: An Open Framework for Unifying and Advancing Discrete Text Optimization

tau

Tel Aviv University

Submitted by

amanmeh

When Agents Commit Too Soon: Diagnosing Premature Commitment in LLM Agents

Snowflake

Submitted by

anmol

Capable but Careless: Do Computer-Use Agents Follow Contextual Integrity?

UKPLab

Ubiquitous Knowledge Processing Lab

Submitted by

hotchpotch

HAKARI-Bench: A Lightweight Benchmark for Comparing Retrieval Architectures and Efficiency Settings under Unified Conditions

hakari-bench

Submitted by

xuyichen

Libretto: Giving LLM Agents a Sense of Musical Structure

UCBerkeley

University of California, Berkeley

Submitted by

ArturoDeza

Robusto-2: Benchmarking Humans & VLMs for Autonomous Driving in Lima & New York City

Artificio

Submitted by

xcjthu

Beyond Reward Engineering: A Data Recipe for Long-Context Reinforcement Learning

openbmb

Submitted by

DDAI-D

AC-ODM: Actor--Critic Online Data Mixing for Sample-Efficient LLM Pretraining

opendatalab

Submitted by

jakubgregorek

ShotcreteDepth: A Bi-modal Dataset for Robust Robotic Depth Perception in Shotcrete Construction Environments

dtu-pcas

DTU - Perception and Cognition for Autonomous Systems

Submitted by

slempert

An Exploratory Case Study of LLM-Assisted Refactoring and Gameplay Feature Generation in an Endless Runner Game

IUInternationalUniversity

IU International University

Submitted by

yonghyunk1m

Improving Text-to-Music Generation with Human Preference Rewards

TuneJury

Submitted by

LoanMaikon

Toward Parking Spot Occupancy Recognition: A Self-Supervised Approach

·
3 authors