arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2501.05855cs.CL
收藏

ConSim:利用自动化可模拟性衡量基于概念的解释的有效性

ConSim: Measuring Concept-Based Explanations' Effectiveness with Automated Simulatability

IRT Saint Exupéry(圣埃克苏佩里技术研究所);IRIT(图卢兹计算机科学研究所);Université Paul-Sabatier(保罗-萨巴蒂埃大学);Google Research(谷歌研究院);CerCo, CNRS, Université de Toulouse(法国国家科学研究中心, 图卢兹大学);ANITI, Université de Toulouse(图卢兹大学)

Antonin Poché, Alon Jacovi, Agustin Martin Picard, Victor Boutin, Fanny Jourdan

AI总结:

本文提出ConSim框架,利用大语言模型作为模拟器,通过自动化可模拟性端到端评估基于概念的解释的有效性,实现跨模型和数据集的可扩展且一致的评估,并验证了LLM排名的一致性。

摘要:

基于概念的解释通过将复杂的模型计算映射到人类可理解的概念来发挥作用。评估此类解释非常困难,因为它不仅包括所诱导的可能概念空间的质量,还包括所选概念如何有效地传达给用户。现有的评估指标往往只关注前者,而忽略了后者。我们引入了一个评估框架,通过自动化可模拟性来衡量概念解释:即模拟器根据所提供的解释预测被解释模型输出的能力。该方法在端到端评估中同时考虑了概念空间及其解释。关于可模拟性的人体研究 notoriously 难以实施,尤其是在广泛、全面的实证评估规模上(这正是本工作的主题)。我们提出使用大语言模型(LLMs)作为模拟器来近似评估,并报告了多种分析以使这种近似可靠。我们的方法允许在各种模型和数据集上进行可扩展且一致的评估。我们使用该框架报告了全面的实证评估,并表明LLMs对解释方法提供了一致的排名。代码可在 https://github.com/AnonymousConSim/ConSim 获取。

英文摘要:

Concept-based explanations work by mapping complex model computations to human-understandable concepts. Evaluating such explanations is very difficult, as it includes not only the quality of the induced space of possible concepts but also how effectively the chosen concepts are communicated to users. Existing evaluation metrics often focus solely on the former, neglecting the latter. We introduce an evaluation framework for measuring concept explanations via automated simulatability: a simulator's ability to predict the explained model's outputs based on the provided explanations. This approach accounts for both the concept space and its interpretation in an end-to-end evaluation. Human studies for simulatability are notoriously difficult to enact, particularly at the scale of a wide, comprehensive empirical evaluation (which is the subject of this work). We propose using large language models (LLMs) as simulators to approximate the evaluation and report various analyses to make such approximations reliable. Our method allows for scalable and consistent evaluation across various models and datasets. We report a comprehensive empirical evaluation using this framework and show that LLMs provide consistent rankings of explanation methods. Code available at https://github.com/AnonymousConSim/ConSim.

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。

↑