Attention Mechanisms Don’t Learn Additive Models: Rethinking Feature Importance for Transformers
Abstract: We address the critical challenge of applying feature attribution methods to the transformer architecture, which dominates current applications in natural language processing and beyond. Traditional attribution methods to explainable AI (XAI) explicitly or implicitly rely on linear or additive surrogate models to quantify the impact of input features on a model's output. In this work, we formally prove an alarming incompatibility: transformers are structurally incapable to align with popular surrogate models for feature attribution, undermining the grounding of these conventional explanation methodologies. To address this discrepancy, we introduce the Softmax-Linked Additive Log-Odds Model (SLALOM), a novel surrogate model specifically designed to align with the transformer framework. SLALOM demonstrates the capacity to deliver a range of insightful explanations with across both synthetic and real-world datasets. We highlight SLALOM's unique efficiency-quality curve by showing that SLALOM can produce explanations with substantially higher fidelity than competing surrogate models or provide explanations of comparable quality at a fraction of their computational costs.
Submission Length: Regular submission (no more than 12 pages of main content)
Changes Since Last Submission: Fix Linebreaks in Title, typo
Assigned Action Editor: ~Shiyu_Chang2
Submission Number: 3531
Loading