EscapeBench：迈向提升语言模型代理的创造性智能
EscapeBench: Towards Advancing Creative Intelligence of Language Model Agents
AI总结：
针对现有基准忽视创造性适应的问题，本文提出密室逃脱基准EscapeBench及框架EscapeAgent，通过前瞻与反思增强创造性推理，实现更高成功率与更少步骤。
摘要：
语言模型代理在长会话规划和推理方面表现出色，但现有基准主要关注具有明确目标的面向目标的任务，忽视了在陌生环境中的创造性适应。为了解决这一问题，我们引入了EscapeBench，一个由密室逃脱游戏环境组成的基准套件，旨在通过创造性推理、非常规工具使用和迭代问题解决来挑战代理，以发现隐含目标。我们的结果表明，当前的语言模型，尽管采用了工作记忆和思维链推理，在没有提示的情况下平均进展仅为15%，凸显了它们在创造力方面的局限性。为了弥合这一差距，我们提出了EscapeAgent，一个旨在通过前瞻性（创新工具使用）和反思（识别未解决任务）来增强创造性推理的框架。实验表明，EscapeAgent能够执行超过1000步的动作链，同时保持逻辑连贯性。它能够以最多减少40%的步骤和提示来导航并完成游戏，在不同难度级别上表现稳健，并实现更高的动作成功率，采用更高效和创新的谜题解决策略。
英文摘要：
Language model agents excel in long-session planning and reasoning, but existing benchmarks primarily focus on goal-oriented tasks with explicit objectives, neglecting creative adaptation in unfamiliar environments. To address this, we introduce EscapeBench, a benchmark suite of room escape game environments designed to challenge agents with creative reasoning, unconventional tool use, and iterative problem-solving to uncover implicit goals. Our results show that current LM models, despite employing working memory and Chain-of-Thought reasoning, achieve only 15% average progress without hints, highlighting their limitations in creativity. To bridge this gap, we propose EscapeAgent, a framework designed to enhance creative reasoning through Foresight (innovative tool use) and Reflection (identifying unsolved tasks). Experiments show that EscapeAgent can execute action chains over 1,000 steps while maintaining logical coherence. It navigates and completes games with up to 40% fewer steps and hints, performs robustly across difficulty levels, and achieves higher action success rates with more efficient and innovative puzzle-solving strategies.
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。
