Explainable AI: Learning from the Learners

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Vinuesa, Ricardo, Brunton, Steven L., Mengaldo, Gianmarco
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917274276855808
author Vinuesa, Ricardo
Brunton, Steven L.
Mengaldo, Gianmarco
author_facet Vinuesa, Ricardo
Brunton, Steven L.
Mengaldo, Gianmarco
contents Artificial intelligence now outperforms humans in several scientific and engineering tasks, yet its internal representations often remain opaque. In this Perspective, we argue that explainable artificial intelligence (XAI), combined with causal reasoning, enables {\it learning from the learners}. Focusing on discovery, optimization and certification, we show how the combination of foundation models and explainability methods allows the extraction of causal mechanisms, guides robust design and control, and supports trust and accountability in high-stakes applications. We discuss challenges in faithfulness, generalization and usability of explanations, and propose XAI as a unifying framework for human-AI collaboration in science and engineering.
format Preprint
id arxiv_https___arxiv_org_abs_2601_05525
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Explainable AI: Learning from the Learners
Vinuesa, Ricardo
Brunton, Steven L.
Mengaldo, Gianmarco
Artificial Intelligence
Machine Learning
Computational Physics
Physics and Society
Artificial intelligence now outperforms humans in several scientific and engineering tasks, yet its internal representations often remain opaque. In this Perspective, we argue that explainable artificial intelligence (XAI), combined with causal reasoning, enables {\it learning from the learners}. Focusing on discovery, optimization and certification, we show how the combination of foundation models and explainability methods allows the extraction of causal mechanisms, guides robust design and control, and supports trust and accountability in high-stakes applications. We discuss challenges in faithfulness, generalization and usability of explanations, and propose XAI as a unifying framework for human-AI collaboration in science and engineering.
title Explainable AI: Learning from the Learners
topic Artificial Intelligence
Machine Learning
Computational Physics
Physics and Society
url https://arxiv.org/abs/2601.05525