大型语言模型概率推理中的双重特质
Dual Traits in Probabilistic Reasoning of Large Language Models
AI总结：
本研究通过三项实验发现大型语言模型在后验概率判断中同时表现出遵循贝叶斯规则的规范模式与依赖相似性的代表性模式，推测其源于强化学习中的对比损失函数，并强调需谨慎部署以减少认知偏差。
摘要：
我们进行了三项实验，以探究大型语言模型（LLMs）如何评估后验概率。我们的结果揭示了最先进模型中后验判断的两种模式共存：一种规范模式，遵循贝叶斯规则；另一种基于代表性的模式，依赖于相似性——这与人类的系统1和系统2思维相似。此外，我们观察到LLMs难以从记忆中回忆基础率信息，并且开发提示工程策略以减轻基于代表性的判断可能具有挑战性。我们进一步推测，判断的双重模式可能是强化学习从人类反馈中采用的对比损失函数的结果。我们的发现强调了减少LLMs认知偏差的潜在方向，以及在关键领域谨慎部署LLMs的必要性。
英文摘要：
We conducted three experiments to investigate how large language models (LLMs) evaluate posterior probabilities. Our results reveal the coexistence of two modes in posterior judgment among state-of-the-art models: a normative mode, which adheres to Bayes' rule, and a representative-based mode, which relies on similarity -- paralleling human System 1 and System 2 thinking. Additionally, we observed that LLMs struggle to recall base rate information from their memory, and developing prompt engineering strategies to mitigate representative-based judgment may be challenging. We further conjecture that the dual modes of judgment may be a result of the contrastive loss function employed in reinforcement learning from human feedback. Our findings underscore the potential direction for reducing cognitive biases in LLMs and the necessity for cautious deployment of LLMs in critical areas.
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。
