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 11

Submitted by

StableKirito

Mean Mode Screaming: Mean--Variance Split Residuals for 1000-Layer Diffusion Transformers

·
1 authors

Submitted by

Osilly

Flow-OPD: On-Policy Distillation for Flow Matching Models

·
11 authors

Submitted by

Jiashuz

MACE-Dance: Motion-Appearance Cascaded Experts for Music-Driven Dance Video Generation

GD-ML

Submitted by

yunqu

Listwise Policy Optimization: Group-based RLVR as Target-Projection on the LLM Response Simplex

Tencent-Hunyuan

Tencent Hunyuan

2

Submitted by

ChengsongHuang

LLMs Improving LLMs: Agentic Discovery for Test-Time Scaling

google

Submitted by

hkuzxc

HyperEyes: Dual-Grained Efficiency-Aware Reinforcement Learning for Parallel Multimodal Search Agents

XiaohongshuAI

Submitted by

taesiri

HumanNet: Scaling Human-centric Video Learning to One Million Hours

·
2 authors

Submitted by

peregrine123

Rubric-based On-policy Distillation

·
10 authors

Submitted by

Yu2020

Anisotropic Modality Align

·
11 authors

Submitted by

Geralt-Targaryen

Beyond Retrieval: A Multitask Benchmark and Model for Code Search

high-quality llm benchmarks

Submitted by

VINHYU

TextLDM: Language Modeling with Continuous Latent Diffusion

JoyFutureAcademy

Joy Future Academy

Submitted by

aldjalkdf

UniPrefill: Universal Long-Context Prefill Acceleration via Block-wise Dynamic Sparsification

tencent

Submitted by

Zhaorun

DecodingTrust-Agent Platform (DTap): A Controllable and Interactive Red-Teaming Platform for AI Agents

Virtue-AI-HUB

Submitted by

jiwan-chung

Rethinking State Tracking in Recurrent Models Through Error Control Dynamics

yonseiworld

Yonsei University

Submitted by

dongdaxiang

AEM: Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning

baidu

Submitted by

ionutmodo

MatryoshkaLoRA: Learning Accurate Hierarchical Low-Rank Representations for LLM Fine-Tuning

ISTA-DASLab

IST Austria Distributed Algorithms and Systems Lab

Submitted by

jankin123

4DThinker: Thinking with 4D Imagery for Dynamic Spatial Understanding

THU1911

Tsinghua University

Submitted by

dxlong2000

A^2RD: Agentic Autoregressive Diffusion for Long Video Consistency

google

Submitted by

fxmeng

MISA: Mixture of Indexer Sparse Attention for Long-Context LLM Inference

PKU

Peking University

Submitted by

griver

Q-RAG: Long Context Multi-step Retrieval via Value-based Embedder Training

·
10 authors

Submitted by

jindongwang

UniSD: Towards a Unified Self-Distillation Framework for Large Language Models

GeorgiaTech

Georgia Institute of Technology

Submitted by

LMMM2025

Scaling Continual Learning to 300+ Tasks with Bi-Level Routing Mixture-of-Experts

TheHKU

Hong Kong University

Submitted by

taesiri

STARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal Generation

apple

Submitted by

taesiri

Normalizing Trajectory Models

apple

Submitted by

YuZeng260

SCOPE: Structured Decomposition and Conditional Skill Orchestration for Complex Image Generation

·
16 authors

Submitted by

evijit

What if AI systems weren't chatbots?

·
4 authors

Submitted by

yuezhengrong

What Matters for Diffusion-Friendly Latent Manifold? Prior-Aligned Autoencoders for Latent Diffusion

alibaba-inc

Submitted by

luisrui

ModelLens: Finding the Best for Your Task from Myriads of Models

ucdavis

Submitted by

yuweiyin

IntentGrasp: A Comprehensive Benchmark for Intent Understanding

UBC-V

University of British Columbia

Submitted by

Fernandez-Owen

SkCC: Portable and Secure Skill Compilation for Cross-Framework LLM Agents

SunYatsen

Sun Yat-Sen University

Submitted by

Bohan22

MLS-Bench: A Holistic and Rigorous Assessment of AI Systems on Building Better AI

·
28 authors

Submitted by

taesiri

InterLV-Search: Benchmarking Interleaved Multimodal Agentic Search

·
8 authors

Submitted by

taesiri

Fast Byte Latent Transformer

·
8 authors

Submitted by

danielhzlin

From Storage to Experience: A Survey on the Evolution of LLM Agent Memory Mechanisms

·
9 authors

Submitted by

danjacobellis

LiVeAction: a Lightweight, Versatile, and Asymmetric Neural Codec Design for Real-time Operation

·
2 authors

Submitted by

huuuuyulong

MDN: Parallelizing Stepwise Momentum for Delta Linear Attention

·
8 authors

Submitted by

wshuai190

DiffRetriever: Parallel Representative Tokens for Retrieval with Diffusion Language Models

ielabgroup

The Information Engineering Lab

Submitted by

shaokaiyeah

BalCapRL: A Balanced Framework for RL-Based MLLM Image Captioning

apple

2

Submitted by

conan1024hao

Sparse Autoencoders as Plug-and-Play Firewalls for Adversarial Attack Detection in VLMs

mtri-admin

Submitted by

weijiezz

SpecBlock: Block-Iterative Speculative Decoding with Dynamic Tree Drafting

·
12 authors

Submitted by

minghaoyan

PACEvolve++: Improving Test-time Learning for Evolutionary Search Agents

google

Submitted by

farukakgul

Rethinking RL for LLM Reasoning: It's Sparse Policy Selection, Not Capability Learning

USC

University of Southern California

Submitted by

shuidi0020

Steering Visual Generation in Unified Multimodal Models with Understanding Supervision

·
8 authors

Submitted by

Arjhun000

Empirical Evidence for Simply Connected Decision Regions in Image Classifiers

UniTuebingen

Eberhard Karls Universität Tübingen

Submitted by

akshat-sj

Discovering Reinforcement Learning Interfaces with Large Language Models

Lossfunk

Submitted by

xyzhang368

Learning Visual Feature-Based World Models via Residual Latent Action

·
6 authors

Submitted by

zsqzz

Who Prices Cognitive Labor in the Age of Agents? Compute-Anchored Wages

UIUC-CS

University of Illinois at Urbana-Champaign

Submitted by

ShinmJS

PrefixGuard: From LLM-Agent Traces to Online Failure-Warning Monitors

TACPS-liv

Trustworthy Autonomous Cyber Physical Systems (ACPS) Lab

Submitted by

Jim137

Gated QKAN-FWP: Scalable Quantum-inspired Sequence Learning

·
19 authors

Submitted by

guosy

CASCADE: Case-Based Continual Adaptation for Large Language Models During Deployment

JilinUniversity001

Jilin University

Submitted by

hadamelino

CGM-JEPA: Learning Consistent Continuous Glucose Monitor Representations via Predictive Self-Supervised Pretraining

google

Submitted by

yeonseokjeong

R^3-SQL: Ranking Reward and Resampling for Text-to-SQL

·
5 authors

Submitted by

JiahaoChen1

From Holo Pockets to Electron Density: GPT-style Drug Design with Density

·
7 authors

Submitted by

cst7

Delta-Adapter: Scalable Exemplar-Based Image Editing with Single-Pair Supervision

·
8 authors

Submitted by

jeongseokoh

Shallow Prefill, Deep Decoding: Efficient Long-Context Inference via Layer-Asymmetric KV Visibility

·
5 authors

Submitted by

wish44165

CPCANet: Deep Unfolding Common Principal Component Analysis for Domain Generalization

unimelb-nlp

The University of Melbourne

Submitted by

aminr8

Trajectory as the Teacher: Few-Step Discrete Flow Matching via Energy-Navigated Distillation

apple