通过反事实反馈激发语言模型的推理能力
Reasoning Elicitation in Language Models via Counterfactual Feedback
AI总结：
针对语言模型反事实因果推理不足的问题，本文提出平衡事实与反事实问答的新指标及多种微调方法，并验证其相对基础模型在归纳、演绎推理任务上的泛化提升。
摘要：
尽管语言模型的有效性不断提升，但其推理能力仍不发达。特别是，通过反事实问答进行因果推理的能力较为欠缺。本工作旨在弥补这一差距。首先，我们推导了新的指标，用于平衡事实性问题和反事实问题的准确率，相比传统仅基于事实的指标，能够更完整地刻画语言模型的推理能力。其次，我们提出了若干微调方法，旨在在所提出指标的意义上激发更好的推理机制。最后，我们在多种现实场景中评估了微调后语言模型的性能。特别是，我们研究了在若干需要归纳和演绎等推理能力的问题中，我们的微调方法在多大程度上相对于基础模型系统性地实现更好的泛化。
英文摘要：
Despite the increasing effectiveness of language models, their reasoning capabilities remain underdeveloped. In particular, causal reasoning through counterfactual question answering is lacking. This work aims to bridge this gap. We first derive novel metrics that balance accuracy in factual and counterfactual questions, capturing a more complete view of the reasoning abilities of language models than traditional factual-only based metrics. Second, we propose several fine-tuning approaches that aim to elicit better reasoning mechanisms, in the sense of the proposed metrics. Finally, we evaluate the performance of the fine-tuned language models in a variety of realistic scenarios. In particular, we investigate to what extent our fine-tuning approaches systemically achieve better generalization with respect to the base models in several problems that require, among others, inductive and deductive reasoning capabilities.
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。
