思考思考本身：SAGE-nano 面向自感知语言模型的逆向推理
Thinking About Thinking: SAGE-nano's Inverse Reasoning for Self-Aware Language Models
AI总结：
本文提出逆向推理范式，使 SAGE-nano 模型能够事后解释自身推理链，在提升推理准确率的同时增强可解释性，并构建了首个 LLM 自我反思与推理透明度评估框架。
摘要：
大型语言模型（LLMs）在利用思维链（Chain-of-Thought, CoT）提示解决复杂推理任务方面已展现出显著能力，但其决策过程在某种程度上仍是一个黑箱。我们引入了逆向推理（inverse reasoning），这是一种新颖的范式，使 LLMs 能够在事后分解并解释自身的推理链。我们的方法用于 SAGE-nano——一个具有 40 亿参数的推理模型，采用了一种元认知结构，通过注意力过程进行回溯，以识别主要决策点并生成对推理选择的解释。典型的 CoT 方法面向正向推理生成，而逆向推理则能揭示为何选择了特定的推理链而非其他推理链。通过对来自 AQUA-RAT、CommonsenseQA 以及定制基准的逻辑推理谜题、数学问题和伦理困境进行全面测试，我们证明 SAGE-nano 在其任务上无论是推理准确率（AQUA-RAT 上为 74.6%）还是解释质量（92.1% 的人类偏好得分）都处于前沿水平，并且其性能几乎可与 Claude-3.5 Sonnet 或 GPT-4o 等模型相媲美。我们的贡献包括：（i）首个通过逆向推理实现 LLM 自我反思的严格框架；（ii）一种用于逆转注意力流的新型元学习框架；（iii）面向推理透明度的综合评估框架；以及（iv）证明利用逆向推理增强推理能够同时提升可解释性和推理性能的证据。我们的工作为透明 AI 系统开辟了新途径，并弥合了 AI 安全、教育和科学发现方面的重大差距。
英文摘要：
Large Language Models (LLMs) have demonstrated remarkable capabilities at solving complex reasoning tasks with Chain-of-Thought (CoT) prompting, but their decision-making processes remain somewhat blackbox. We introduce textbfinverse reasoning, a novel paradigm enabling LLMs to decompose and explain their own reasoning chains post-hoc. Our approach, used in SAGE-nano, a 4-billion-parameter reasoning model, employs a metacognitive structure that reflects back via attention processes to identify major decision points and generate explanations of reasoning choices. While typical CoT approaches are directed towards forward reasoning generation, inverse reasoning provides insight into why specific reasoning chains were selected over others. Through thorough testing of logical reasoning puzzles, math problems and ethical dilemmas from AQUA-RAT, CommonsenseQA, and customized benchmarks, we demonstrate that SAGE-nano is at the cutting edge both on reasoning accuracy (74.6% on AQUA-RAT) and explanation quality (92.1% human preference score) for its task, and offers performance almost on par with models like Claude-3.5 Sonnet or GPT-4o. Our contributions are: (i) the first rigorous framework for LLM self-reflection via inverse reasoning, (ii) a novel metalearning framework to reverse the attention flow, (iii) comprehensive evaluation frameworks for reasoning transparency, and (iv) evidence that increasing reasoning using inverse reasoning improves interpretability along with reasoning performance. Our work creates new avenues for transparent AI systems and closes significant gaps in AI safety, education, and scientific discovery.
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。
