大语言模型生成任务中解释的反事实可模拟性
Counterfactual Simulatability of LLM Explanations for Generation Tasks
AI总结：
本文提出了一种评估大语言模型生成任务中解释反事实可模拟性的通用框架，发现其在新闻摘要任务中表现良好，但在医疗建议任务中仍有改进空间。
摘要：
大语言模型（LLMs）可以是不可预测的，因为即使对提示进行微小的修改，也会导致输出以意想不到的方式变化。因此，模型准确解释其行为的能力至关重要，尤其是在高风险场景中。评估解释的一种方法是反事实可模拟性，即解释在多大程度上允许用户推断模型在相关反事实情况下的输出。反事实可模拟性之前已被研究用于是/否问答任务。我们提供了一个通用框架，用于将这种方法扩展到生成任务，使用新闻摘要和医疗建议作为示例用例。我们发现，尽管LLM解释在摘要设置中使用户能够更好地预测LLM在反事实情况下的输出，但在医疗建议方面仍有显著改进空间。此外，我们的结果表明，反事实可模拟性的评估可能更适合技能型任务，而不是知识型任务。
英文摘要：
LLMs can be unpredictable, as even slight alterations to the prompt can cause the output to change in unexpected ways. Thus, the ability of models to accurately explain their behavior is critical, especially in high-stakes settings. One approach for evaluating explanations is counterfactual simulatability, how well an explanation allows users to infer the model's output on related counterfactuals. Counterfactual simulatability has been previously studied for yes/no question answering tasks. We provide a general framework for extending this method to generation tasks, using news summarization and medical suggestion as example use cases. We find that while LLM explanations do enable users to better predict LLM outputs on counterfactuals in the summarization setting, there is significant room for improvement for medical suggestion. Furthermore, our results suggest that the evaluation for counterfactual simulatability may be more appropriate for skill-based tasks as opposed to knowledge-based tasks.
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。
