Tag
SERA（Self-Evaluating Recursive Agents）提出一个联合训练单一策略以实现子任务分解、执行与评估的框架，通过可验证结果对齐评估准则并引入叶覆盖奖励，显著优于递归智能体基线，表明学会评估能提升求解能力。
COPE is a novel optimization framework for continual personalization of large language models under sparse user feedback, using learnable user embeddings and self-evaluation calibration. Experiments show it outperforms training-free and training-based baselines and remains robust in various settings.
This paper identifies the 'progress mirage' failure mode in long-running autonomous LLM agents, where self-evaluation bias causes agents to mistake stagnation for progress. Through controlled experiments, it shows that external, out-of-band verification is necessary for open-ended objectives.
A comprehensive survey of 1,250 papers (2024–2026) on recursive self-improvement in AI, proposing a taxonomy distinguishing bounded self-refinement from open-ended recursive self-improvement, and analyzing the evaluator design space and failure modes.
This paper tests the assumption that LLMs judge better than they generate in in-context QA, finding generation accuracy exceeds self-evaluation on most benchmarks, with evaluation attending less to context. The findings challenge core assumptions in self-evaluation pipelines.
This paper introduces Self-Evaluation Elicitation (SEE), which uses calibration-coupled reinforcement learning and masked distillation to elicit latent judge calibration in base LLMs with minimal data, improving calibration across benchmarks while preserving answer quality.