人工智能中的递归自我改进：从有界自我精炼到自主研究循环

摘要

人工智能系统正日益参与自身的改进：修订输出结果、在部署过程中调整自身适配器、基于自身生成的数据进行训练，并越来越广泛地开展人工智能研究本身。现有文献在描述这一现象时使用的词汇（如“自我精炼”“自我奖励”“自我博弈”“自我进化”）混淆了几种根本上不同的目标。我们调研了2024-2026年间1,250篇arXiv论文，沿两个维度展开分析：系统改进的对象——是部署中的行为、通过训练获得的策略、评估器，还是研究过程本身；以及闭环的程度（从人在回路中到完全闭环）。该分类法将有界自我精炼——收敛、可评估且已成为工业实践——与开放式递归自我改进（RSI）区分开来，后者在每一个可测量的维度上都仍受限于立足需求、崩溃动力学和计算约束。其独特之处在于设立了专门的“自我评估”类别：每一个改进回路都隐含一个主张，即某种信号可以替代人类判断。我们调研了评估器的设计空间——包括评判模型、过程奖励模型、验证器、评分标准、元评估——将信号按从形式验证器（最强）到内在自我评估（最弱）的顺序排列成一个验证层次结构，并观察到演示的自我改进强度与该层次结构一致，其失效模式（自我确证循环、模型崩溃、多样性崩溃）源于对该层次结构的违反，而将人类留在回路中的“研究方向设定瓶颈”则位于该层次结构的顶端。我们将技术文献与关于RSI限制的理论、以及关于闭环问题中前沿实验室报告所带来的安全与治理问题联系起来，并指出在治理层面测量自我改进能力是该领域最缺乏研究的空白地带。

English

AI systems increasingly participate in their own improvement: revising their outputs, adapting their own harnesses during deployment, training on data they generate, and, increasingly, conducting AI research itself. This literature is described under a vocabulary ("self-refine," "self-reward," "self-play," "self-evolve") that conflates fundamentally different ambitions. We survey 1,250 arXiv papers (2024-2026) along two axes: what the system improves -- its behavior in deployment, its policy through training, its evaluator, or the research process itself -- and the degree of loop closure (human-in-the-loop to fully closed). The taxonomy separates bounded self-refinement -- convergent, evaluable, and already industrial practice -- from open-ended recursive self-improvement (RSI), which remains bounded by grounding requirements, collapse dynamics, and compute constraints on every measured axis. Its distinctive feature is a dedicated category for self-evaluation: every improvement loop is a claim that some signal can substitute for human judgment. We survey the evaluator design space -- judges, process reward models, verifiers, rubrics, meta-evaluation -- order the signals into a verification hierarchy from formal verifiers (strongest) to intrinsic self-assessment (weakest), and observe that demonstrated self-improvement strength tracks this hierarchy, that its failure modes (self-confirming loops, model collapse, diversity collapse) follow from its violations, and that the "research direction-setting" bottleneck keeping humans in the loop sits at the top of that hierarchy. We connect the technical literature to the theory of RSI limits and to the safety and governance questions raised by frontier-lab accounts of closing the loop, and identify governance-grade measurement of self-improvement as the field's most underpopulated niche.

人工智能中的递归自我改进：从有界自我精炼到自主研究循环

Recursive Self-Improvement in AI: From Bounded Self-Refinement to Autonomous Research Loops

摘要

Support