A Close Look at Decomposition-based XAI-Methods for Transformer Language Models

TL;DR

This paper provides a comprehensive evaluation of decomposition-based XAI methods for transformer language models, comparing ALTI-Logit, LRP, and AttnLRP, and introduces a benchmark dataset for future research.

Contribution

It conducts the first detailed quantitative and qualitative comparison of prominent decomposition-based XAI methods on language models, and releases a benchmark dataset and code.

Findings

ALTI-Logit and LRP show differing strengths in attribution accuracy.

AttnLRP offers a promising extension with improved interpretability.

Gradient-based methods provide complementary insights.

Abstract

Various XAI attribution methods have been recently proposed for the transformer architecture, allowing for insights into the decision-making process of large language models by assigning importance scores to input tokens and intermediate representations. One class of methods that seems very promising in this direction includes decomposition-based approaches, i.e., XAI-methods that redistribute the model's prediction logit through the network, as this value is directly related to the prediction. In the previous literature we note though that two prominent methods of this category, namely ALTI-Logit and LRP, have not yet been analyzed in juxtaposition and hence we propose to close this gap by conducting a careful quantitative evaluation w.r.t. ground truth annotations on a subject-verb agreement task, as well as various qualitative inspections, using BERT, GPT-2 and LLaMA-3 as a testbed. Along the way we compare and extend the ALTI-Logit and LRP methods, including the recently proposed AttnLRP variant, from an algorithmic and implementation perspective. We further incorporate in our benchmark two widely-used gradient-based attribution techniques. Finally, we make our carefullly constructed benchmark dataset for evaluating attributions on language models, as well as our code, publicly available in order to foster evaluation of XAI-methods on a well-defined common ground.