AI summaryⓘ
The authors created S³Gym, a tool to test if language models can learn from interacting with games by trying things out (Self-Testing), judging how well they did (Self-Judging), and then improving based on that (Self-Improvement). They tested three ways the models could use their experience: remembering history directly, summarizing important parts, and retraining their parameters. Their results show that improvement depends on the task and method; summaries help when broad rules apply, while raw history is better when details matter. Also, retraining parameters helps sometimes but can also cause problems. This work helps understand how language models can better learn from experience rather than just repeat fixed behaviors.
Large Language ModelsSelf-ImprovementSelf-TestingSelf-JudgingInteractive BenchmarkIn-Context Learning (ICL)Summary MemoryParameter TrainingText-Based GamesBehavioral Experience
Authors
Jiajun Shi, Siyuan Tao, Yuhao Wu, Zexuan Wang, Jingyuan Zhang, Jiaheng Liu, Xinping Lei, Xinrong Zhang, Siyuan Fang, Zhewen Tan, Tianle Cai, Junhao Fang, Jiameng Huang, Yueyang Wang, Jinkai Liu, Yuxuan Zhang, Jian Yang, Zhoujun Li, Shen Yan, Wenhao Huang, Ge Zhang
Abstract
Large language models (LLMs) increasingly interact with external environments and accumulate substantial behavioral experience, yet existing agent benchmarks largely evaluate them as fixed policies. It therefore remains unclear whether an agent can actively test its behavior, judge the resulting experience, and use that experience to improve future decisions. We introduce \textbf{S\textsuperscript{3}Gym}, an interactive benchmark for evaluating LLM self-improvement through three coupled capabilities: \textbf{Self-Testing}, \textbf{Self-Judging}, and \textbf{Self-Improvement}. S$^3$Gym separates permissive exploration from strict held-out evaluation and instantiates this protocol in seven text-based games with executable environment verifiers. We evaluate three pathways for incorporating interaction experience: direct History ICL, score-conditioned Summary Memory, and parameter Training. Our experiments reveal that self-improvement is neither automatic nor uniform. Context-level experience improves performance for several model--game pairs, but the most effective pathway depends strongly on the task structure: summaries are beneficial when experience can be compressed into reusable strategic rules, yet often underperform raw history when success depends on precise, state-contingent information. Parameter training produces substantial gains on some tasks, but also exhibits unstable improvement and severe negative transfer on others. These findings show that recognizing successful actions is insufficient; agents must also transform feedback into executable and transferable policies. S$^3$Gym provides a unified framework for diagnosing this process and identifying the bottlenecks that prevent agents from translating interaction experience into reliable self-improvement.