Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2512.02735 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866918228352040960 |
|---|---|
| author | Bjøru, Anna Rodum Lysnæs-Larsen, Jacob Jørgensen, Oskar Strümke, Inga Langseth, Helge |
| author_facet | Bjøru, Anna Rodum Lysnæs-Larsen, Jacob Jørgensen, Oskar Strümke, Inga Langseth, Helge |
| contents | This work presents a conceptual framework for causal concept-based post-hoc Explainable Artificial Intelligence (XAI), based on the requirements that explanations for non-interpretable models should be understandable as well as faithful to the model being explained. Local and global explanations are generated by calculating the probability of sufficiency of concept interventions. Example explanations are presented, generated with a proof-of-concept model made to explain classifiers trained on the CelebA dataset. Understandability is demonstrated through a clear concept-based vocabulary, subject to an implicit causal interpretation. Fidelity is addressed by highlighting important framework assumptions, stressing that the context of explanation interpretation must align with the context of explanation generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_02735 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Framework for Causal Concept-based Model Explanations Bjøru, Anna Rodum Lysnæs-Larsen, Jacob Jørgensen, Oskar Strümke, Inga Langseth, Helge Artificial Intelligence This work presents a conceptual framework for causal concept-based post-hoc Explainable Artificial Intelligence (XAI), based on the requirements that explanations for non-interpretable models should be understandable as well as faithful to the model being explained. Local and global explanations are generated by calculating the probability of sufficiency of concept interventions. Example explanations are presented, generated with a proof-of-concept model made to explain classifiers trained on the CelebA dataset. Understandability is demonstrated through a clear concept-based vocabulary, subject to an implicit causal interpretation. Fidelity is addressed by highlighting important framework assumptions, stressing that the context of explanation interpretation must align with the context of explanation generation. |
| title | A Framework for Causal Concept-based Model Explanations |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2512.02735 |