Explore · 每周精选
发现最受关注的研究论文,追踪研究趋势,订阅感兴趣的期刊与关键词。
50 篇论文 · 按点赞排序
Deyuan Liu, Yihao Hu, Jingxuan Zhang +13 authors
UltraText Bench evaluates dense visual text generation across bilingual scenes using structured references and a vision-language judge, revealing trade-offs between clarity, fidelity, and workload difficulty.
Wen Xue, Yong Liu, Xueheng Li +5 authors
LayerEditor unifies instruction-guided image editing with reusable RGBA layer generation via joint token modeling and edit-prior distillation, enabling precise local manipulation and layer reuse.
Jihai Zhang, Runjia Qian, Kaiqin Kong +15 authors
Matrix-Game 4.0 separates game logic from visual synthesis via an executable engine and autoregressive video renderer, achieving real-time interactive gameplay with low latency.
Xiaomin Yu, Hongliang Chen, Zhe Huang +22 authors
OpenDS-World introduces a live-web benchmark to evaluate autonomous multi-step deep search agents, revealing significant gaps between frontier models and human performance.
Jiaming Zhang, Xuan Wang, Fuyao Zhang +3 authors
A decoder-free, action-conditioned world model predicts next-screen representations to verify GUI transitions and detect hijacked screens with high accuracy and low latency.
Xinhao Ji, Hui En Pang, Hao Li +12 authors
EgoXperience introduces a unified framework for joint reconstruction of camera motion, scene geometry, hands, and objects from egocentric video using automatically annotated large-scale data and a single-pass model.
Haoxiang Zhang, Qinglin Chen, Hiroaki Hayashi +9 authors
Prospective learning via Self-Retrospection Distillation uses hindsight experience to train foresight predictions, improving reinforcement learning when outcome rewards are uniform.
Zhiyong Cao, Hengjie Zhu, Yongxin Wang +20 authors
TraceTrial-Bench evaluates agent harness reliability across seven components using mechanism-oriented tasks, trajectory tracing, and failure attribution to compare harness designs and diagnose failures.
Hao Li, MeiJia Chen, Weijie Ren +4 authors
RIDE improves on-policy distillation by extrapolating reinforcement-learning-induced representation shifts directly in hidden states, yielding stable gains over RL-trained teachers.
Ruhao Liu, Qi Li, Shizun Wang +1 authors
DynaFlow enables adaptive multi-agent collaboration by letting a main agent synthesize executable workflows that coordinate sub-agents through structured execution and verification.
Team Kandinsky, Julia Agafonova, Bulat Akhmatov +85 authors
Kandinsky 6.0 Video is a family of diffusion models that generate synchronized audio-video clips with lip-sync via a dual-stream CrossDiT architecture and multi-stage training.
Ruopeng Gao, Yukun Huang, Jiarui Hu +4 authors
Future Chunk Prediction improves autoregressive video diffusion by enforcing future-chunk supervision, yielding more physically plausible video generation with reduced training cost.
Hao Li, Xiyuan Zheng, Yuanyuan Gao +7 authors
HOLI-WORLD is an integrated ecosystem that automates the construction of large-scale learning resources from egocentric recordings to improve machine understanding of dynamic scenes and interactions.
Zhaoxi Zhang, Hao Liang, Linxuan Wu +7 authors
SpatialScout evaluates vision-language agents on active 3D object mapping, revealing major gaps in precise metric map construction.
Pan Wang, Yihao Hu, Guangyuan Dong +8 authors
Progress-Aligned Recurrence trains looped language models to match intermediate states and stabilize recurrent norms, enabling strong length extrapolation on sequential tracking tasks.
Mingze Sun, Ziang Cao, Yuan Fei +8 authors
DONIC is a generalist bimanual tracker trained on diverse human-motion data that uses reinforcement learning with recycled demonstrations and per-hand rewards to achieve high success on unseen objects without fine-tuning.
Tianyu Fu, Ruoxi Wang, Zihan Min +6 authors
AKV introduces cache engineering to manage LLM agent key-value caches without altering trajectories, improving efficiency and accuracy via drop and repos operations with a co-designed serving engine.
Seungone Kim, Jiayi Geng, Young-Jun Lee +25 authors
Researchers benchmarked AI agents on open scientific problems by having them draft follow-up papers from unpublished directions, finding that none met expert standards and most contained fabricated claims or contradictory evidence.
Panrong Tong, Quyu Kong, Pengxiang Li +12 authors
MAI-Gym is an open platform that unifies GUI agent evaluation, data collection, and online reinforcement learning training across mobile, computer, and web environments, enabling scalable synthesis of verified tasks and training of a unified 9B agent with state-of-the-art open performance.
Bingchen Yao, Haobo Xu, Haokun Lin +6 authors
STEPQuant is a spatial-temporal quantization framework that reduces memory usage of linear attention recurrent states by allocating precision based on error magnitude and memory lifetime.
北京市昌平区探索星信息技术及软件开发工作室
京ICP备2026059466号