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 4

Submitted by

taesiri

UniVidX: A Unified Multimodal Framework for Versatile Video Generation via Diffusion Priors

·
11 authors

Submitted by

Zhouhc

Web2BigTable: A Bi-Level Multi-Agent LLM System for Internet-Scale Information Search and Extraction

·
9 authors

Submitted by

taesiri

Map2World: Segment Map Conditioned Text to 3D World Generation

·
5 authors

Submitted by

etaisella

Prox-E: Fine-Grained 3D Shape Editing via Primitive-Based Abstractions

·
6 authors

Submitted by

unknowncloudw

From Skill Text to Skill Structure: The Scheduling-Structural-Logical Representation for Agent Skills

PekingUniversity

Peking University

3

Submitted by

jianlanluo

Learning while Deploying: Fleet-Scale Reinforcement Learning for Generalist Robot Policies

·
16 authors

2

Submitted by

taesiri

Let ViT Speak: Generative Language-Image Pre-training

ByteDance

Submitted by

calintz

Stable-GFlowNet: Toward Diverse and Robust LLM Red-Teaming via Contrastive Trajectory Balance

·
6 authors

Submitted by

milkkarten

Odysseus: Scaling VLMs to 100+ Turn Decision-Making in Games via Reinforcement Learning

PrincetonUniversity

Princeton University

Submitted by

rsn243

Trees to Flows and Back: Unifying Decision Trees and Diffusion Models

TUM

Technical University of Munich

1

Submitted by

eugene6923

When Do Diffusion Models learn to Generate Multiple Objects?

·
5 authors

Submitted by

taesiri

End-to-End Autoregressive Image Generation with 1D Semantic Tokenizer

ByteDance-Seed

Submitted by

iieycx

Online Self-Calibration Against Hallucination in Vision-Language Models

·
6 authors

Submitted by

rajkumarrawal

Learning to Act and Cooperate for Distributed Black-Box Consensus Optimization

·
4 authors

Submitted by

iNeil77

Themis: Training Robust Multilingual Code Reward Models for Flexible Multi-Criteria Scoring

project-themis

Submitted by

Jona-te-Lintelo

MASCing: Configurable Mixture-of-Experts Behavior via Activation Steering Masks

·
5 authors

Submitted by

vijaysadashivaiah

Better Models, Faster Training: Sigmoid Attention for single-cell Foundation Models

Merck

Merck & Co., Inc.

Submitted by

tobiaslee

AnalogRetriever: Learning Cross-Modal Representations for Analog Circuit Retrieval

·
5 authors

1

Submitted by

praxelhq

LASE: Language-Adversarial Speaker Encoding for Indic Cross-Script Identity Preservation

Praxel

Submitted by

danielhzlin

Talker-T2AV: Joint Talking Audio-Video Generation with Autoregressive Diffusion Modeling

·
11 authors

Submitted by

frankzydou

Soft Anisotropic Diagrams for Differentiable Image Representation

·
3 authors