通过递归自我批评实现超人类AI的可扩展监督
Scalable Oversight for Superhuman AI via Recursive Self-Critiquing
AI总结：
本文提出通过递归自我批评实现超人类AI的可扩展监督，探索批评的批评比直接批评更容易，并验证递归批评在AI监督中的有效性。
摘要：
随着AI能力在复杂任务中日益超越人类 proficiency，当前的对齐技术，包括SFT和RLHF，在确保可靠监督方面面临根本性挑战。这些方法依赖直接的人类评估，当AI输出超过人类认知阈值时变得不切实际。为应对这一挑战，我们探讨了两个假设：(1) 对批评的批评比批评本身更容易，将广泛接受的观察扩展到批评领域，因为批评本身就是一种专门的生成；(2) 这种困难关系具有递归性，建议当直接评估不可行时，执行更高阶的批评（例如批评的批评的批评）提供了一条更可行的监督途径。我们进行了人-人、人-AI和AI-AI实验，以研究递归自我批评对AI监督的潜力。我们的结果强调递归批评作为可扩展AI监督的有前途的方法。
英文摘要：
As AI capabilities increasingly surpass human proficiency in complex tasks, current alignment techniques, including SFT and RLHF, face fundamental challenges in ensuring reliable oversight. These methods rely on direct human assessment and become impractical when AI outputs exceed human cognitive thresholds. In response to this challenge, we explore two hypotheses: (1) \textit{Critique of critique can be easier than critique itself}, extending the widely-accepted observation that verification is easier than generation to the critique domain, as critique itself is a specialized form of generation; (2) \textit{This difficulty relationship holds recursively}, suggesting that when direct evaluation is infeasible, performing higher-order critiques (e.g., critique of critique of critique) offers a more tractable supervision pathway. We conduct Human-Human, Human-AI, and AI-AI experiments to investigate the potential of recursive self-critiquing for AI supervision. Our results highlight recursive critique as a promising approach for scalable AI oversight.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
