Tag
推文介绍了一个关于递归自我改进(RSI)的精选研究合集,涵盖从STaR、Self-Rewarding LMs到Promptbreeder和Automated Design of Agentic Systems等关键论文,梳理了RSI从自生成训练到自我修改代码代理的演进脉络。
A personal collection of 1,700 Arxiv whitepapers organized into 90 categories with cross-linking and synthesis, now publicly available as a web tool called Inquiring Lines.