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 2

Submitted by

EddieYang428

Adaptive Reward Routing: Dynamic Multi-Reward Optimization for Joint Audio-Video Diffusion via Forward-Process RL

tencent

Submitted by

rhfeiyang

Hierarchical Continuous Diffusion Language Models

UIUC-CS

University of Illinois at Urbana-Champaign

Submitted by

liushiliushi

Agent Priors-guided Policy Learning

NationalUniversityofSingapore

National University of Singapore

Submitted by

MingLiiii

A Missing Piece for Trustworthy AI Reviewers: From Benchmarking Rhetorical Robustness to SciCore Review

·
7 authors

Submitted by

IDKTreeDP

Beyond Memory: Harnessing Long-Horizon Agents with Explicit Belief States

alibabagroup

Submitted by

enrue1893

World Observer: Joint Actor-Observer Generation for Persistent World Modeling

kaist-ai

Submitted by

changdae

Sharpening Tax in Post-Training

meta

Submitted by

pshlego

ActiveSaddler: Automated Curriculum Learning for Agent Harness Optimization

microsoft

Submitted by

YangC777

AutoGUIWorld: Image Generators as Visual World Models for GUI Agent

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

wangzx1994

ROWBench: Do Video Models Render What the Program Specifies?

AlayaLab

Submitted by

allonsy07

Retrieval-Augmented Skill Optimization via Cross-Harness Adaptation

·
11 authors

Submitted by

groundhogLLM

GraphForge: Training Working Agents with Graph-Anchored Workspace Synthesis

ustc-community

University of Science and Technology of China

Submitted by

b1o1o1m

Beyond the Current Scene: Event-Referential Grasping with Active View Selection

Seoul National University

Submitted by

la0ka1

Scaling and Distilling Text Embeddings for Better Diffusibility

umich

University of Michigan

Submitted by

zhaihaotian

Make Sparse Rewards Count: Density-Aware Reward Aggregation for Multi-Reward RL

kuaishou technology

Submitted by

vveicao

4Director: Controlling Video World Models with Rigid 3D Geometry

stabilityai

Submitted by

chuxuan

RPTune: Learned Context Curation for LLM Catalog Search

google

Submitted by

owl10

Multimodal Flow: Unified Flow Modeling of Language and Vision in Embedding Spaces

HUST Vision Lab

Submitted by

gowitheflow

AutoDataBench: A Data-centric Testbed for Accelerating Auto Research

AutoDataBench

Submitted by

isminoula

SILSA: Sliding-Window Slice Latents for Topology-Preserving High-Resolution 3D Generation

PLAN-Lab

PLAN Lab @University of Illinois Urbana-Champaign

Submitted by

xusirui

InterEvolve: Test-Time Evolution of Reward Programs for Humanoid Loco-Manipulation

UIUC-CS

University of Illinois at Urbana-Champaign

Submitted by

Yushi98

When Users Change Their Minds: Measuring and Repairing Intent Drift in LLM Agents

·
4 authors

Submitted by

Yushi98

Do Audio LLMs Listen Before They Act? Diagnosing Acoustic-Context Gating in Voice Agents

·
3 authors

Submitted by

xsourse

PixelDense: Dense Prediction as Representation Alignment for Pixel Diffusion

·
12 authors

Submitted by

Luffuly

Memorizon: Training World Models Beyond Their Context Window

MBZUAI-IFM

Submitted by

nanamma

Benchmarking and Enhancing Skill-Level Memory for Partially Observable Robotic Manipulation

·
7 authors

Submitted by

xypkent

DataMagic: Authoring Data Videos through Declarative Multi-Agent Orchestration

·
6 authors

Submitted by

luisrui

Smaller Models, Better Rejects: Preference Distillation Scaling

Submitted by

yunuyean

Prefill-Free Cross-Family KV Cache Transfer for Heterogeneous Multi-Agent LLMs

UniversityofSouthernCalifornia

University of Southern California

Submitted by

gtomitsu

Argo-Bench: Evaluating Data Agents on Enterprise-Scale Workflows

textql

Submitted by

Franck-Dernoncourt

FlexRouter: Learning Complementary Model Sets for Flexible LLM Routing

·
9 authors

Submitted by

Franck-Dernoncourt

Controlled Decoding Attacks on Black-Box LLMs

·
7 authors

Submitted by

Franck-Dernoncourt

Joint and Cross-Modal Video-Audio Generation and Editing: A Unified Formulation and Design Taxonomy

·
24 authors

1

Submitted by

WHB139426

OmniSeek: Native Tool Integration for Multi-turn Audio-Visual Reasoning

·
8 authors

Submitted by

lr10260

Align Then Reason: A Multimodal Lip-Sync Judge for Dubbing

netflix

Submitted by

ethanning

Predictive Credit: Measuring What Scientific Explanations Add to Experimental Forecasts

CarnegieMellonU

Carnegie Mellon University

Submitted by

ethanning

Rules to Tools: Executable Checks for LLM Agents in Scientific Computing

AbhiVakil29112001

Carnegie Mellon University

Submitted by

jiangnanhugo

Explore Broadly, Reason Sharply: Push Small Models toward the Frontier via Sampling

morganstanley

Submitted by

wjs0725

Does Native 3D Texture Generation Necessarily Require 3D Assets for Training?

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

xuehang

Pay for the Fault, Not the Flow: Label-Free In-Flow Multi-Agent Workflow Optimization

·
7 authors

Submitted by

Franck-Dernoncourt

Personalized Image Generation with Reasoning and Reflection

·
13 authors

Submitted by

Franck-Dernoncourt

Prompt2Skill: Unsupervised Skill Optimization From Natural Language Instructions

·
5 authors

Submitted by

AIGNLAI

Pretrain Once, Route Anywhere: Towards a Foundation Model for LLM Routing

nanjinguniv

Nanjing University

Submitted by

SteveZeyuZhang

DexPolicy: Scheduled Exploration for Trajectory-Guided Dexterous Manipulation

PekingUniversity

Peking University

Submitted by

1202kbs

Learning What to Recall: Adaptive Multi-Cue Episodic Memory for World Models

kaist-ai

Submitted by

xuehang

It Takes Workflows to Evolve Better Workflows

·
6 authors