The Hydra Effect: Emergent Self-repair in Language Model ComputationsThomas McGrath,Matthew Rahtz,Janos Kramar,Vladimir Mikulik,Shane LeggCoRR(2023)引用 23|浏览130关键词language,hydra effect,model,self-repairAI 理解论文溯源树样例生成溯源树,研究论文发展脉络Chat Paper正在生成论文摘要