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

Feb 9

Submitted by

Myashka

F-GRPO: Don't Let Your Policy Learn the Obvious and Forget the Rare

t-tech

2

Submitted by

fairyang

Baichuan-M3: Modeling Clinical Inquiry for Reliable Medical Decision-Making

baichuan-inc

Baichuan Intelligent Technology

Submitted by

xufangzhi

OdysseyArena: Benchmarking Large Language Models For Long-Horizon, Active and Inductive Interactions

·
19 authors

Submitted by

Kushnareva

AudioSAE: Towards Understanding of Audio-Processing Models with Sparse AutoEncoders

huawei-noah

HUAWEI Noah's Ark Lab

Submitted by

Mystery

On the Entropy Dynamics in Reinforcement Fine-Tuning of Large Language Models

·
7 authors

Submitted by

rmndrnts

Pisets: A Robust Speech Recognition System for Lectures and Interviews

NSU

Novosibirsk State University

Submitted by

lx865712528

MSign: An Optimizer Preventing Training Instability in Large Language Models via Stable Rank Restoration

microsoft

3

Submitted by

taesiri

DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos

nvidia

Submitted by

yfqiu-nlp

Self-Improving World Modelling with Latent Actions

·
7 authors

Submitted by

amphora

Judging What We Cannot Solve: A Consequence-Based Approach for Oracle-Free Evaluation of Research-Level Math

·
8 authors

Submitted by

taesiri

POINTS-GUI-G: GUI-Grounding Journey

·
10 authors

Submitted by

master-lan

Self-Improving Multilingual Long Reasoning via Translation-Reasoning Integrated Training

·
9 authors

3

Submitted by

ly4096

Canzona: A Unified, Asynchronous, and Load-Balanced Framework for Distributed Matrix-based Optimizers

Qwen

Qwen

Submitted by

LiPengyi29

Back to Basics: Revisiting Exploration in Reinforcement Learning for LLM Reasoning via Generative Probabilities

·
4 authors

Submitted by

lgy0404

MemGUI-Bench: Benchmarking Memory of Mobile GUI Agents in Dynamic Environments

zju

Zhejiang University

Submitted by

jankin123

OmniVideo-R1: Reinforcing Audio-visual Reasoning with Query Intention and Modality Attention

tencent

Submitted by

taesiri

EgoAVU: Egocentric Audio-Visual Understanding

·
12 authors

Submitted by

taesiri

InftyThink+: Effective and Efficient Infinite-Horizon Reasoning via Reinforcement Learning

zju

Zhejiang University

Submitted by

JingzeShi

OmniMoE: An Efficient MoE by Orchestrating Atomic Experts at Scale

BAAI

Beijing Academy of Artificial Intelligence

Submitted by

p-song1

Large Language Model Reasoning Failures

StanfordUniversity

Stanford University

Submitted by

d7chong

RaBiT: Residual-Aware Binarization Training for Accurate and Efficient LLMs

SamsungResearch

Samsung Research

3

Submitted by

xw-eric

Group-Evolving Agents: Open-Ended Self-Improvement via Experience Sharing

·
6 authors

3

Submitted by

refrain-wbh

Outcome Accuracy is Not Enough: Aligning the Reasoning Process of Reward Models

Qwen

Qwen

Submitted by

ylu610

Uncovering Cross-Objective Interference in Multi-Objective Alignment

DM2-ND

University of Notre Dame DM2 Lab

Submitted by

comparIA

compar:IA: The French Government's LLM arena to collect French-language human prompts and preference data

ministere-culture

Ministère de la Culture (SNUM)

Submitted by

nanzhang

QuantLRM: Quantization of Large Reasoning Models via Fine-Tuning Signals

Pennsylvania State University

Submitted by

fmmarkmq

SEMA: Simple yet Effective Learning for Multi-Turn Jailbreak Attacks

microsoft

Submitted by

Duke-de-Artois

PlanViz: Evaluating Planning-Oriented Image Generation and Editing for Computer-Use Tasks

·
10 authors

Submitted by

taesiri

SeeUPO: Sequence-Level Agentic-RL with Convergence Guarantees

Tongyi-MAI

Submitted by

DIYIN

ReMiT: RL-Guided Mid-Training for Iterative LLM Evolution

tencent

3

Submitted by

ambroiseodt

Vision Transformer Finetuning Benefits from Non-Smooth Components

huawei-noah

HUAWEI Noah's Ark Lab

Submitted by

FengTing

Revisiting the Shape Convention of Transformer Language Models

MediaTek-Research

MediaTek Research

2

Submitted by

dayll

SEAD: Self-Evolving Agent for Multi-Turn Service Dialogue

meituan

Submitted by

g-luo

Learning a Generative Meta-Model of LLM Activations

generative-latent-prior

Generative Latent Prior

3

Submitted by

GMFTBY

Table-as-Search: Formulate Long-Horizon Agentic Information Seeking as Table Completion

AIDC-AI

Submitted by

hhua2

SPARC: Separating Perception And Reasoning Circuits for Test-time Scaling of VLMs

ibm-research

2

Submitted by

stan-hua

Uncertainty Drives Social Bias Changes in Quantized Large Language Models

·
3 authors

2

Submitted by

olaflaitinen

Urban Spatio-Temporal Foundation Models for Climate-Resilient Housing: Scaling Diffusion Transformers for Disaster Risk Prediction

·
3 authors

2

Submitted by

akhaliq

SE-Bench: Benchmarking Self-Evolution with Knowledge Internalization

·
6 authors

Submitted by

liangtm

Seg-ReSearch: Segmentation with Interleaved Reasoning and External Search

iSEE-Laboratory

iSEE-Laboratory

Submitted by

yousefkotp

AtlasPatch: An Efficient and Scalable Tool for Whole Slide Image Preprocessing in Computational Pathology

AtlasAnalyticsLab

Atlas Analytics Lab

2

Submitted by

Jinyang23

Exploring Knowledge Purification in Multi-Teacher Knowledge Distillation for LLMs

·
8 authors

3

Submitted by

leon0923

Avoiding Premature Collapse: Adaptive Annealing for Entropy-Regularized Structural Inference

·
1 authors

3