The Quarrel of Local Post-hoc Explainers for Moral Values Classification in Natural Language Processing
by Andrea Agiollo, Luciano C. Siebert, Pradeep K. Murukannaiah and Andrea Omicini
Although popular and effective, large language models (LLM) are characterised by a performance vs. transparency trade-off that hinders their applicability to sensitive scenarios. This is the main reason behind many approaches focusing on local post-hoc explanations recently proposed by the XAI community. However, to the best of our knowledge, a thorough comparison among available explainability techniques is currently missing, mainly for the lack of a general metric to measure their benefits. We compare state-of-the-art local post-hoc explanation mechanisms for models trained over moral value classification tasks based on a measure of correlation. By relying on a novel framework for comparing global impact scores, our experiments show how most local post-hoc explainers are loosely correlated, and highlight huge discrepancies in their results—their “quarrel” about explanations. Finally, we compare the impact scores distribution obtained from each local post-hoc explainer with human-made dictionaries, and point out that there is no correlation between explanation outputs and the concepts humans consider as salient.
- DOI: https://doi.org/10.1007/978-3-031-40878-6_6
- URL: https://link.springer.com/content/pdf/10.1007/978-3-031-40878-6_6.pdf
@incollection{quarrel-extraamas2023,
apice = {QuarrelExtraamas2023},
author = {Agiollo, Andrea and Siebert, Luciano C. and Murukannaiah, Pradeep K. and Omicini, Andrea},
booktitle = {Explainable and Transparent {AI} and Multi-Agent Systems},
chapter = 6,
dblp = {conf/extraamas/AgiolloSMO23},
doi = {10.1007/978-3-031-40878-6_6},
editor = {Calvaresi, Davide and Najjar, Amro and Omicini, Andrea and Aydoǧan, Reyhan and Carli, Rachele and Ciatto, Giovanni and Mualla, Yazan and Främling, Kary},
iris = {11585/940734},
isbn = {978-3-031-40878-6},
issn = {0302-9743},
keywords = {Natural Language Processing, Moral Values Classification, eXplainable Artificial Intelligence, Local Post-hoc Explanations.},
month = sep,
numpages = 19,
pages = {97--115},
publisher = {Springer},
scholar = {16243619767560534499},
scopus = {2-s2.0-85172189313},
series = {Lecture Notes in Computer Science},
subseries = {Lecture Notes in Artificial Intelligence},
title = {The Quarrel of Local Post-hoc Explainers for Moral Values Classification in Natural Language Processing},
url = {http://link.springer.com/10.1007/978-3-031-40878-6_6},
urlpdf = {https://link.springer.com/content/pdf/10.1007/978-3-031-40878-6_6.pdf},
volume = 14127,
year = 2023
}