迈向对大语言模型中行为引导干预的可靠评估
Towards Reliable Evaluation of Behavior Steering Interventions in LLMs
AI总结：
本文针对大语言模型行为引导干预评估缺乏客观量化标准的问题，提出包含四项准则的可靠评估流程，并据此发现部分表示工程干预方法的实际效果低于此前报告。
摘要：
表示工程方法近期在实现高效的模型行为引导方面展现出前景。然而，这些方法的评估流程主要依赖主观演示，而非定量、客观的指标。我们旨在通过倡导当前评估中缺失的四项属性来着手解决这一问题：（i）应使用与下游任务足够相似的上下文来评估干预质量；（ii）应考虑模型似然度；（iii）评估应允许在不同目标行为之间进行标准化比较；（iv）应提供基线比较。我们引入了一个基于这些准则的评估流程，提供定量和可视化分析，以衡量给定方法的实际效果。我们使用该流程评估两种表示工程方法在引导真实性和可修正性等行为方面的有效性，发现某些干预措施的效果不如先前报告的那样显著。
英文摘要：
Representation engineering methods have recently shown promise for enabling efficient steering of model behavior. However, evaluation pipelines for these methods have primarily relied on subjective demonstrations, instead of quantitative, objective metrics. We aim to take a step towards addressing this issue by advocating for four properties missing from current evaluations: (i) contexts sufficiently similar to downstream tasks should be used for assessing intervention quality; (ii) model likelihoods should be accounted for; (iii) evaluations should allow for standardized comparisons across different target behaviors; and (iv) baseline comparisons should be offered. We introduce an evaluation pipeline grounded in these criteria, offering both a quantitative and visual analysis of how effectively a given method works. We use this pipeline to evaluate two representation engineering methods on how effectively they can steer behaviors such as truthfulness and corrigibility, finding that some interventions are less effective than previously reported.
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。
