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
Haolin Chen, Deon Metelski, Leon Qi +30 authors
Healthcare workflow benchmark challenges agents with policy-dense, multi-role, and multilateral interaction requirements, revealing significant performance gaps in automated enterprise applications.
Chengzhi Liu, Yichen Guo, Yepeng Liu +8 authors
LLM agents executing within execution harnesses can produce correct outputs while violating safety constraints during execution, necessitating trajectory-level auditing to ensure proper resource access and information flow across multi-agent systems.
Hanwen Wang, Weizhi Zhao, Xiangyu Wang +11 authors
DexJoCo presents a benchmark and toolkit for dexterous manipulation with 11 functional tasks evaluating tool-use, bimanual coordination, and long-horizon execution, along with a low-cost data collection system and comprehensive model evaluation.
Jinyuan Li, Langlin Huang, Chengsong Huang +5 authors
BetaPRM introduces a distributional approach to process reward models that predicts both success probabilities and prediction reliability, enabling adaptive computation allocation that reduces token usage while maintaining accuracy.
Minrui Xu, Zilin Wang, Mengyi DENG +12 authors
EnvFactory automates the creation of executable tool environments and natural multi-turn trajectories for training LLMs with agentic reinforcement learning, achieving superior performance with fewer resources.
Zhepei Wei, Xinyu Zhu, Wei-Lin Chen +3 authors
Reinforcement learning with verifiable rewards parameter trajectories exhibit low-rank structures that enable efficient extrapolation through a simple linear regression method, demonstrating superior performance with reduced computational requirements.
Yifan Dai, Zhenhua Wu, Bohan Zeng +18 authors
LatentOmni is a cross-modal reasoning framework that interleaves textual reasoning with audio-visual latent states using feature-level supervision and temporal consistency embedding, outperforming explicit text-based chain-of-thought approaches in audio-visual reasoning tasks.
Ziliang Zhao, Zenan Xu, Shuting Wang +7 authors
PlanningBench introduces a framework for generating scalable and verifiable planning data that enables better evaluation and training of large language models' planning capabilities through structured taxonomies and constraint-driven synthesis.
Jung Yi, Minjae Kim, Paul Hyunbin Cho +3 authors
WorldKV enables persistent world generation in video diffusion models by retrieving and compressing key-value cache chunks to maintain consistency while improving throughput.
Javad Rajabi, Kimia Shaban, Koorosh Roohi +2 authors
SEGA improves high-resolution text-to-image generation by adaptively scaling attention across RoPE components based on spatial-frequency structure during denoising steps.
Taewon Yun, Jisu Shin, Jeonghwan Choi +2 authors
CoRD is a collaborative multi-teacher decoding framework that synthesizes reasoning trajectories through predictive perplexity scoring and beam search, enabling efficient distillation of large reasoning models with high-quality outputs and generalized performance.
Yixuan Yang, Zhen Luo, Wanshui Gan +5 authors
A novel MLLM-based agentic framework called Code-as-Room generates 3D indoor rooms by converting top-down images into executable Blender code through a structured execution harness with cross-stage memory to maintain context.
Emaad Khwaja, Chris Lettieri, Gerald Woo +10 authors
Time series foundation models demonstrate scalable forecasting performance across parameter sizes, with Toto 2.0 achieving state-of-the-art results on multiple benchmarks through a unified training approach.
Zunhai Su, Rui Yang, Chao Zhang +11 authors
OScaR is a novel KV cache compression framework that addresses token norm imbalance through canalized rotation and omni-token scaling, achieving significant improvements in memory efficiency and decoding speed for extended context language models.
Ruicheng Zhang, Kaixi Cong, Jun Zhou +5 authors
ODENative online GRPO framework KVPO aligns streaming video generators with human preferences through causal-semantic exploration and velocity-field surrogate policy based on trajectory velocity energy.
Banghao Chi, Yining Xie, Mingyuan Wu +9 authors
Spreadsheet-RL is a reinforcement learning framework that trains specialized spreadsheet agents in realistic Excel environments, improving AI agent performance on both general and domain-specific spreadsheet tasks through automated data collection and domain-specific benchmarks.
Hongjun Liu, Yifei Ming, Shafiq Joty +1 authors
HASP introduces executable program functions that serve as active guardrails for LLM agents, enabling direct intervention in agent loops and improving performance across complex tasks.
Xiaoxuan He, Siming Fu, Zeyue Xue +9 authors
Flash-GRPO improves training efficiency for video diffusion models by addressing temporal variance and gradient inconsistency through iso-temporal grouping and temporal gradient rectification.
Hongji Yang, Songlian Li, Yucheng Zhou +4 authors
Diffusion models applied in compressed image space generate high-quality images with lower computational cost and support flexible inputs like text or boxes.
Yang Yue, Fangyun Wei, Tianyu He +10 authors
InsightTok improves discrete visual tokenization for better text and face reconstruction through content-aware perceptual losses, enhancing autoregressive image generation quality.
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