arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.11252cs.CL
收藏

反事实作为评估自回归语言模型中归因方法忠实性的手段

Counterfactuals As a Means for Evaluating Faithfulness of Attribution Methods in Autoregressive Language Models

Tehran Institute for Advanced Studies, Khatam University(哈塔姆大学德黑兰高级研究所)

Sepehr Kamahi, Yadollah Yaghoobzadeh

AI总结:

针对自回归语言模型归因方法忠实性评估中破坏输入导致分布外样本的问题,本文提出利用反事实生成技术构造流畅且符合分布的反事实样本,以提高评估协议的可靠性。

摘要:

尽管自回归语言模型已被广泛采用,可解释性评估研究却主要集中在跨度填充和掩码语言模型上。评估一种解释方法的忠实性——即它在多大程度上准确解释模型的内部工作机制和决策过程——具有挑战性,因为很难将模型与其解释分离开来。大多数忠实性评估技术会破坏或移除被特定归因(特征重要性)方法判定为重要的输入标记,并观察模型输出随之发生的变化。然而,对于自回归语言模型而言,由于其下一标记预测的训练目标,这种方法会产生分布外输入。在本研究中,我们提出一种利用反事实生成来评估自回归语言模型归因方法忠实性的技术。我们的技术能够生成流畅且符合分布的反事实样本,从而使评估协议更加可靠。

英文摘要:

Despite the widespread adoption of autoregressive language models, explainability evaluation research has predominantly focused on span infilling and masked language models. Evaluating the faithfulness of an explanation method -- how accurately it explains the inner workings and decision-making of the model -- is challenging because it is difficult to separate the model from its explanation. Most faithfulness evaluation techniques corrupt or remove input tokens deemed important by a particular attribution (feature importance) method and observe the resulting change in the model's output. However, for autoregressive language models, this approach creates out-of-distribution inputs due to their next-token prediction training objective. In this study, we propose a technique that leverages counterfactual generation to evaluate the faithfulness of attribution methods for autoregressive language models. Our technique generates fluent, in-distribution counterfactuals, making the evaluation protocol more reliable.

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。

↑