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 3

Submitted by

namespace-ERI

Repo-To-Skill: Distilling GitHub Repositories Into AI4AI Skills

BAAI

Beijing Academy of Artificial Intelligence

Submitted by

mozhu

HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness?

ByteDance-Seed

Submitted by

mozhu

Aspire: Can Models Self-Evolve from Vague Goals?

ByteDance-Seed

Submitted by

junchao-cuhk

SolarWM: Open Data and Scalable Training for Long-Horizon Video World Models

CUHKSZ

Chinese University of Hong Kong, Shenzhen

Submitted by

YerbaPage

EarlyEval: Cheaper Agent Evaluation via Early Outcome Prediction

SJTU

Shanghai Jiao Tong University

Submitted by

itsnamgyu

Language Models Can Control Their Own Attention

kaist-ai

Submitted by

Leo-Dai

It Takes Two to Match: Co-Evolving Generative Retriever with Reinforcement Learning

apple

2

Submitted by

mozhu

S3Gym: Can LLMs Turn Self-Testing and Self-Judging into Self-Improvement?

ByteDance-Seed

Submitted by

HaeChan0305

WHALE: A Simple Recipe for Joint Harness-Weight Optimization

KRAFTON

Submitted by

tonywu71

NeoMME: A Single-Tower Multimodal-Native Multilingual Foundation Encoder for Efficient Fine-Tuning and Inference

Hcompany

2

Submitted by

gowitheflow

On the Design Fundamentals of Pixel Text Representation Learning

LCO-Embedding

Submitted by

Hangrui-Xu

Beyond Visual Similarity: Entity-Aligned Retrieval for Knowledge-Based Visual Question Answering

Tsinghua University

2

Submitted by

lhmd

ZipTok3D: High-Fidelity 3D Tokenization with Compact Token Prefixes

zju

Zhejiang University

2

Submitted by

HakHan

Cliff: Learning Process Rewards from the First Mistake

·
6 authors

Submitted by

Franck-Dernoncourt

A Glance Is All You Need: Single-Pass Fine-Grained Image Captioning with SimLoss

adobe-research

Submitted by

zzliang

VibeVoice-ASR-Streaming Technical Report

microsoft

Submitted by

Leo-Dai

Influence-Directed Distillation: Solving the Diversity Bottleneck in Sampled-Token On-Policy Distillation

Leander001

2

Submitted by

taesiri

Post-Training Language Models for Gold-Medal Performance in Coding Competitions

nvidia

Submitted by

MatteoCargnelutti

Institutional Newspapers Pipeline: Deriving billions of high quality tokens from historical newspapers

institutional

Institutional Data Initiative

2

Submitted by

seokwon99

MULTI3IR: A Benchmark for Multi-perspective Multi-domain Multi-modal Information Retrieval

SeoulNatlUniv

Seoul National University

2

Submitted by

Franck-Dernoncourt

CRISP: Cliff-awaRe Input-adaptive Sparse Prefilling with Structural-Mass-Motivated Routing

·
7 authors

Submitted by

AaryanK

ExecRetrieval: Measuring the Functional-Correctness Gap in Code-Embedding Retrieval

kennesaw

Kennesaw State University

2

Submitted by

jaehong31

PaperCompiler: Faithful Paper-to-Code Generation via Repository-Level Specification Compilation

nanyang-technological-university-singapore

Nanyang Technological University Singapore

Submitted by

ssingh22

Exploring Collaboration between a language and a non-language agent

adobe

Submitted by

zrchen03

SnapBench: Benchmarking Snap-and-Ask Multimodal Retrieval for Mobile Interactions

·
10 authors

3

Submitted by

zachtian

Ignorance or Incompetence? Constructing Knowledge-Gated, Verifiable Tasks for LLM Agents

·
8 authors

4

Submitted by

ashnedungadi

Autoregressive Mosaics: Probing 2D Spatial Reasoning in Text-Only Language Models

·
3 authors

2

Submitted by

taesiri

Kirin: Animal Motion Generation from In-the-Wild Video

·
5 authors

Submitted by

yeonseokjeong

Replacing Training with Memory: Listwise Selection for Text-to-SQL

snu

Seoul National University

Submitted by

hematteo

Sparse Readout Prism: Explaining Logit-Lens Scores in Features Instead of Tokens

·
4 authors

Submitted by

Ignotus6043

Small Language Models as Judges for Rubric-Based Reinforcement Learning

·
5 authors

2

Submitted by

iproskurina

Debias-SparseGPT: Bias-Aware Pruning for Large Language Models

·
4 authors

3

Submitted by

kate-feingold

FoldingAgent: Inferring Parametric Origami Procedures from Demonstration Videos

·
4 authors

Submitted by

marcusinthesky

Portfolio Risk Bounds without Cross-Asset Return Covariances: Distributional Fields from Language-Model Representations

uct

University of Cape Town

2

Submitted by

marcusinthesky

Wasserstein-Barycentric Interaction Fields for Spatial Factor Models: Evidence from Language-Model Representations

uct

University of Cape Town

2

Submitted by

rohan2810

An Empirical Study on Zero-Data Bootstrapping for Conversational Recommender Systems

UCSanDiego

University of California at San Diego