Quantifying the Overlap: Attribution Maps and Linguistic Heuristics in Encoder-Decoder Machine Translation Models

Aria Nourbakhsh, Salima Lamsiyah, Christoph Schommer


Abstract
Explainable AI (XAI) attribution methods seek to illuminate the decision-making process of generative models by quantifying the contribution of each input token to the generated output. Different attribution algorithms, often rooted in distinct methodological frameworks, can produce varied interpretations of feature importance. In this study, we utilize attribution mappings derived from three distinct methods as weighting signals during the training of encoder-decoder models. Our findings demonstrate that Attention and Value Zeroing attribution weights consistently lead to improved model performance. To better understand the linguistic information these mappings capture, we extract part-of-speech (POS), dependency, and named entity recognition (NER) tags from the input-output pairs and compare them with the XAI attribution maps. Although the Saliency method shows greater alignment with POS and dependency annotations than Value Zeroing, it exhibits more divergence in places where its attributions do not conform to these linguistic tags, compared to the other two methods, and it contributes less to the models’ performance.
Anthology ID:
2025.ranlp-1.94
Volume:
Proceedings of the 15th International Conference on Recent Advances in Natural Language Processing - Natural Language Processing in the Generative AI Era
Month:
September
Year:
2025
Address:
Varna, Bulgaria
Editors:
Galia Angelova, Maria Kunilovskaya, Marie Escribe, Ruslan Mitkov
Venue:
RANLP
SIG:
Publisher:
INCOMA Ltd., Shoumen, Bulgaria
Note:
Pages:
822–831
Language:
URL:
https://aclanthology.org/2025.ranlp-1.94/
DOI:
Bibkey:
Cite (ACL):
Aria Nourbakhsh, Salima Lamsiyah, and Christoph Schommer. 2025. Quantifying the Overlap: Attribution Maps and Linguistic Heuristics in Encoder-Decoder Machine Translation Models. In Proceedings of the 15th International Conference on Recent Advances in Natural Language Processing - Natural Language Processing in the Generative AI Era, pages 822–831, Varna, Bulgaria. INCOMA Ltd., Shoumen, Bulgaria.
Cite (Informal):
Quantifying the Overlap: Attribution Maps and Linguistic Heuristics in Encoder-Decoder Machine Translation Models (Nourbakhsh et al., RANLP 2025)
Copy Citation:
PDF:
https://aclanthology.org/2025.ranlp-1.94.pdf