dair-ai 发布本周精选 AI 论文清单(11 月 16-22 日)
这是 dair-ai 每周整理的论文清单,一口气列出 7 篇,AutoRL 和 EfficientSkill 讲强化学习,Safe Personal RSI 和另外两篇都围绕模型自我改进,还专门讨论了安全性约束。做研究或追论文的人直接当周报目录用,省得自己刷 arXiv。
dair-ai 于 11 月 16-22 日发布了本周 AI 论文精选,共收录 7 篇,包括 AutoRL、EfficientSkill、Safe Personal RSI、CompoundingKernel、Self-Improvement Memory、Weight-Space Programming 和 Self-Directed Self-Improvement。
从主题分布看,本周清单明显集中在两个方向:一是强化学习与技能学习,AutoRL 和 EfficientSkill 都属于这一类,分别针对自动化强化学习流程和高效技能获取;二是自我改进与记忆机制,Safe Personal RSI、Self-Improvement Memory、Self-Directed Self-Improvement 三篇论文都在探索模型如何安全地进行自我迭代,其中 Safe Personal RSI 明确把安全性作为约束条件。
剩余两篇偏向基础方法层面:CompoundingKernel 关注核方法,Weight-Space Programming 则从权重空间的角度提出编程化思路。对研究者来说,这类周度清单可以用来快速跟踪 RSI(递归自我改进)相关论文的脉络,此前 dair-ai 每周都会以类似格式汇总,读者可在原帖中查看每篇论文的详细解读。
The Top AI Papers of the Week (November 16 - 22):
- AutoRL - EfficientSkill - Safe Personal RSI - CompoundingKernel - Self-Improvement Memory - Weight-Space Programming - Self-Directed Self-Improvement
Read on for more:
This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include AutoRL, reinforcement learning, RSI, paper highlights.
Source attribution: curated from DAIR.AI (@dair_ai).