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

Oct 21

Submitted by

zhangshaolei

DeepAnalyze: Agentic Large Language Models for Autonomous Data Science

RUC-DataLab

Submitted by

CCCCCC

Glyph: Scaling Context Windows via Visual-Text Compression

·
14 authors

Submitted by

Andrew613

PICABench: How Far Are We from Physically Realistic Image Editing?

·
13 authors

Submitted by

taesiri

FineVision: Open Data Is All You Need

huggingface

Submitted by

taesiri

RL makes MLLMs see better than SFT

naver-ai

Submitted by

VLyb

TrajSelector: Harnessing Latent Representations for Efficient and Effective Best-of-N in Large Reasoning Model

ZGCA

Zhongguancun Academy

3

Submitted by

SnowNation

Towards Mixed-Modal Retrieval for Universal Retrieval-Augmented Generation

RUC

Renmin University of China

Submitted by

jvamvas

QueST: Incentivizing LLMs to Generate Difficult Problems

microsoft

3

Submitted by

yoon6503

When to Ensemble: Identifying Token-Level Points for Stable and Fast LLM Ensembling

kaist-ai

3

Submitted by

EiffL

AION-1: Omnimodal Foundation Model for Astronomical Sciences

polymathic-ai

Submitted by

Vfrz

Annotation-Efficient Universal Honesty Alignment

·
7 authors

Submitted by

monurcan

Visual Autoregressive Models Beat Diffusion Models on Inference Time Scaling

·
3 authors

2

Submitted by

chestnutlzj

Uniworld-V2: Reinforce Image Editing with Diffusion Negative-aware Finetuning and MLLM Implicit Feedback

PekingUniversity

Peking University

Submitted by

zachary-yin

ConsistEdit: Highly Consistent and Precise Training-free Visual Editing

·
4 authors

Submitted by

Ningyu

Executable Knowledge Graphs for Replicating AI Research

antgroup

Submitted by

abdulfatir

Chronos-2: From Univariate to Universal Forecasting

amazon

Submitted by

shun-zheng

Deep Self-Evolving Reasoning

microsoft

2

Submitted by

sdzy

Beyond Pipelines: A Survey of the Paradigm Shift toward Model-Native Agentic AI

BJTUniversity

Beijing JiaoTong University

Submitted by

taesiri

Embody 3D: A Large-scale Multimodal Motion and Behavior Dataset

metaresearch

Submitted by

taesiri

Enterprise Deep Research: Steerable Multi-Agent Deep Research for Enterprise Analytics

Salesforce

Submitted by

para-lost

Constantly Improving Image Models Need Constantly Improving Benchmarks

UCBerkeley

University of California, Berkeley

Submitted by

taesiri

UltraCUA: A Foundation Model for Computer Use Agents with Hybrid Action

apple

Submitted by

yy0514

Agentic Reinforcement Learning for Search is Unsafe

UniOxford

University of Oxford

Submitted by

passing2961

MultiVerse: A Multi-Turn Conversation Benchmark for Evaluating Large Vision and Language Models

Submitted by

NaiveUser

What Limits Agentic Systems Efficiency?

Wisconsin

University of Wisconsin - Madison

2

Submitted by

xwjzds

Distractor Injection Attacks on Large Reasoning Models: Characterization and Defense

AmazonScience

Submitted by

hongyuyang23casia

Knowledge-based Visual Question Answer with Multimodal Processing, Retrieval and Filtering

CASIA

Chinese Academic of Science Institute of Automation

2

Submitted by

austinxu87

Foundational Automatic Evaluators: Scaling Multi-Task Generative Evaluator Training for Reasoning-Centric Domains

Salesforce

Submitted by

shuaichenchang

Automated Composition of Agents: A Knapsack Approach for Agentic Component Selection

·
8 authors

Submitted by

sayandsarkar

GuideFlow3D: Optimization-Guided Rectified Flow For Appearance Transfer

gradient-spaces

Gradient Spaces Research Group

2

Submitted by

jacksukk

On Non-interactive Evaluation of Animal Communication Translators

·
3 authors

Submitted by

monurcan

Balanced Multi-Task Attention for Satellite Image Classification: A Systematic Approach to Achieving 97.23% Accuracy on EuroSAT Without Pre-Training

·
1 authors

Submitted by

sanskxr02

Beacon: Single-Turn Diagnosis and Mitigation of Latent Sycophancy in Large Language Models

·
4 authors

2

Submitted by

linyueqian

AsyncVoice Agent: Real-Time Explanation for LLM Planning and Reasoning

·
7 authors

2

Submitted by

kellycyy

MoReBench: Evaluating Procedural and Pluralistic Moral Reasoning in Language Models, More than Outcomes

·
18 authors

Submitted by

Zihao-Li

Test-Time Scaling of Reasoning Models for Machine Translation

Helsinki-NLP

Language Technology Research Group at the University of Helsinki

2