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

Aug 26

Submitted by

supermodelteam

GigaBrain-0.7: Scaling Embodied Foundation Models to Emergent Capabilities with a Three-System Architecture

open-gigaai

Submitted by

lyhisme

Annotations as Rollouts: Efficient and Scalable Reinforcement Learning for Video MLLMs

·
7 authors

Submitted by

JUNJIE99

WeMM-Embedding: WeChat Multi-Modal Embedding Technical Report

tencent

Submitted by

pshlego

AutoSaddler: Automatic Harness Optimization with Durable Updates from Agent Execution Traces

microsoft

Submitted by

WeiChow

On-Policy Self-Distillation in Diffusion Models

ByteDance-Seed

Submitted by

pybbb

SecOPD: Mitigating Adaptive Prompt Injections by On-Policy Distillation

·
4 authors

Submitted by

csjiaya

CyberFactory: Scaling Cyber Security Capabilities with Instances from the Wild

IQuestLab

Submitted by

Lingaaaaaaa

Recursive Experiential-Working Memory Evolution for Long-Horizon Agent Harnesses

princetonu

Princeton University

Submitted by

taesiri

On-policy Distillation with Verifiable Reward

·
8 authors

Submitted by

QPHutu

Best Practice Critic Optimization

NUS-DSA3101

National University of Singapore

Submitted by

shuaishuaicdp

Game2World Engine: Unlocking In-the-Wild Gameplay Videos for World Model Training

GameWorldData

Submitted by

HaojunChen

From Seeing to Acting: Smart Glasses as First-Person Intelligence Platforms

·
3 authors

Submitted by

taesiri

Meta^n: Recursive Self-Improvement through Emergent Depth

·
4 authors

Submitted by

taesiri

LAION-BVD: A 10-Million-Hour Open Video Dataset for Multimodal Pre-training

·
12 authors

Submitted by

taesiri

CAFE: Self-Improving Search Agents Need Co-Evolving Feedback

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

seonglae

AgentRoom: Concurrent Multi-Agent Coding in a CRDT-Backed Shared Workspace

·
2 authors

1

Submitted by

seonglae

Automata from Agent Traces: Failure and Next-Step Prediction

holistic-ai

1

Submitted by

YanZhanPKU

Length-Adaptive Decoding for Masked Diffusion Machine Translation

PekingUniversity

Peking University

Submitted by

dev23user

DREAM Technical Report

alibabagroup

1

Submitted by

BanmaXM

When "Must" Becomes "Maybe": Constraint Weakening in LLM Agent Workflows

·
6 authors

1

Submitted by

pyxploiter

MoTE: Mixture of Task Experts for Multi-Task Video Understanding

·
4 authors

Submitted by

kaizhao

TorchMorph: CUDA-accelerated Morphological Transforms

intcomp

Intelligent Computing Laboratory

Submitted by

stephenchungmh

Autonomous Mathematical Discovery in an Open-World Multi-Agent Environment

·
3 authors

Submitted by

FortyTwoo

Latent Action as Intention Enables Efficient Future Imagination for World Action Models

·
13 authors

1