技术雷达TECH RADAR · GITHUB · ARXIV · HACKER NEWS

Agent skills package specialized knowledge and workflows into reusable resources that extend AI agent capabilities. Recent work automatically discovers such skills from agent experience, which enables

Liyan Tang, Cyrus Rashtchian, Chun-Sung Ferng

To improve large language models' ability to resolve real-world software issues, prior work has focused on constructing large-scale agent trajectory datasets and performing supervised fine-tuning (SFT

Dewu Zheng, Ruizhe Ye, Yanlin Wang

In real-world software development, code review typically involves iterative interactions between developers and reviewers to improve software quality, making the process costly and time-consuming. Al

Dewu Zheng, Yanlin Wang, Xiwen Wang

LLM-based agents are increasingly deployed in product-level execution harnesses, where jailbreaks can trigger harmful tool use and persistent state changes, creating greater risks than unsafe text gen

Junjie Zhang, Hui Liu, Kecheng Chen

Chemical reactions are fundamentally transformations in electron space, yet most machine learning approaches model them either through \textit{de novo} generation of product molecules or through heuri

Nguyen Xuan-Vu, Octavian Susanu, Daniel Armstrong

Large language model (LLM) agents in governed organizations must let the persona (instructions, tone, self-presentation) evolve freely, while keeping execution (stateful, audited work) traceable. A si

Yisen Xi

Static scanners are increasingly used to identify executable or otherwise unsafe content in machine- learning artifacts, yet conventional evaluation metrics characterize only cases where a scanner yie

Qianlong Lan, Vinothini Pandurangan, Anuj Kaul

Currently used sepsis severity indices rely on fixed variables and weights established decades ago, which are coarsely discretized and calibrated to a cohort that no longer reflects contemporary criti

Kevin Zhu, Ryan Zhang, Baraa Abed

Friend recommendation is inherently graph-structured: the relevance of a potential connection depends on multi-hop social context rather than user attributes alone. However, deploying message-passing

Maksim Utushkin, Andrei Ovsiannikov, Alexander D'yakonov

State-of-the-art action-conditioned video models are typically restricted to a single robot embodiment, preventing them from leveraging the vast corpus of heterogeneous video data that contains rich s

Kechen Liu, Ola Shorinwa

原文 · 讨论

原文 · 讨论

原文 · 讨论

原文 · 讨论

原文 · 讨论

原文 · 讨论

原文 · 讨论

原文 · 讨论

原文 · 讨论

原文 · 讨论