"Faithful to What?" On the Limits of Fidelity-Based Explanations
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914487283482624 |
|---|---|
| author | Eshbaugh, Jackson |
| author_facet | Eshbaugh, Jackson |
| contents | In explainable AI, surrogate models are commonly evaluated by their fidelity to a neural network's predictions. Fidelity, however, measures alignment to a learned model rather than alignment to the data-generating signal underlying the task. This work introduces the linearity score $λ(f)$, a diagnostic that quantifies the extent to which a regression network's input--output behavior is linearly decodable. $λ(f)$ is defined as an $R^2$ measure of surrogate fit to the network. Across synthetic and real-world regression datasets, we find that surrogates can achieve high fidelity to a neural network while failing to recover the predictive gains that distinguish the network from simpler models. In several cases, high-fidelity surrogates underperform even linear baselines trained directly on the data. These results demonstrate that explaining a model's behavior is not equivalent to explaining the task-relevant structure of the data, highlighting a limitation of fidelity-based explanations when used to reason about predictive performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_12176 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | "Faithful to What?" On the Limits of Fidelity-Based Explanations Eshbaugh, Jackson Machine Learning In explainable AI, surrogate models are commonly evaluated by their fidelity to a neural network's predictions. Fidelity, however, measures alignment to a learned model rather than alignment to the data-generating signal underlying the task. This work introduces the linearity score $λ(f)$, a diagnostic that quantifies the extent to which a regression network's input--output behavior is linearly decodable. $λ(f)$ is defined as an $R^2$ measure of surrogate fit to the network. Across synthetic and real-world regression datasets, we find that surrogates can achieve high fidelity to a neural network while failing to recover the predictive gains that distinguish the network from simpler models. In several cases, high-fidelity surrogates underperform even linear baselines trained directly on the data. These results demonstrate that explaining a model's behavior is not equivalent to explaining the task-relevant structure of the data, highlighting a limitation of fidelity-based explanations when used to reason about predictive performance. |
| title | "Faithful to What?" On the Limits of Fidelity-Based Explanations |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2506.12176 |