CEBaB: Estimating the Causal Effects of Real-World Concepts on NLP Model Behavior

TL;DR

This paper introduces CEBaB, a benchmark dataset for evaluating concept-based explanations in NLP by analyzing how real-world concepts influence model outputs through causal inference.

Contribution

We propose CEBaB, a novel dataset with annotated counterfactual reviews, enabling systematic assessment of concept-based explanation methods in NLP.

Findings

CEBaB allows for quantitative comparison of explanation methods.

Different explanation techniques vary significantly in effectiveness.

The dataset facilitates understanding of how abstract concepts affect model behavior.

Abstract

The increasing size and complexity of modern ML systems has improved their predictive capabilities but made their behavior harder to explain. Many techniques for model explanation have been developed in response, but we lack clear criteria for assessing these techniques. In this paper, we cast model explanation as the causal inference problem of estimating causal effects of real-world concepts on the output behavior of ML models given actual input data. We introduce CEBaB, a new benchmark dataset for assessing concept-based explanation methods in Natural Language Processing (NLP). CEBaB consists of short restaurant reviews with human-generated counterfactual reviews in which an aspect (food, noise, ambiance, service) of the dining experience was modified. Original and counterfactual reviews are annotated with multiply-validated sentiment ratings at the aspect-level and review-level. The rich structure of CEBaB allows us to go beyond input features to study the effects of abstract, real-world concepts on model behavior. We use CEBaB to compare the quality of a range of concept-based explanation methods covering different assumptions and conceptions of the problem, and we seek to establish natural metrics for comparative assessments of these methods.