xWhyL: Causal Interactive Learning
Abstract
Explanations are central to causal reasoning, and cognitive science has long established that the human drive to explain is itself a mechanism for learning about causality. Despite this, learning from those abductive signals is largely ignored in artificial intelligence. While explainable AI (XAI) increasingly draws on causal models to generate explanations, the converse direction about what explanations can do for causality remains largely unexplored. To fill this gap, we propose xWhyL, a formal framework connecting causality and XAI by learning causal models from explanations. We develop a mathematical theory that translates explanations into a learning signal complementary to observational data, and demonstrate how it enables overcoming the limits of observational causal discovery. As explanations can be derived from incorrect beliefs and clash with data, a tension we call the Causal Tug-of-War, we prove conditions under which our framework rejects misspecified explanations rather than absorbing them. Our practical instantiation, Causal Interactive Learning (CIL), shows how expert explanations can efficiently support causal discovery and distinguish correct from incorrect explanations.
Related
- Causal Retention in Interactive Agents: Interface Factorization and Selective Adaptation
- Interactive Visualization for Geometry Learning: The Impact of Augmented Reality on Students' Critical Thinking Skills
- IncVNM: An Interactive Recommender System with Incremental von Neumann-Morgenstern Utility Learning
- Harnessing human expertise for high-precision robotic assembly in industrialized construction: A sample-efficient installer-in-the-loop interactive reinforcement learning framework
- FOCUS: Fixed-Confidence Online Causal Learning Using Sequential Adaptive Interventions
BibTeX
@misc{tagliapietra2026xwhyl,
title = {{xWhyL: Causal Interactive Learning}},
author = {Nicholas Tagliapietra and Florian Peter Busch and Moritz Willig and Matej Zečević and Lavdim Halilaj and Juergen Luettin and Kristian Kersting},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.26037},
eprint = {2609.26037},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.26037},
note = {CC BY-NC-SA 4.0},
}
RIS
TY - UNPB
TI - xWhyL: Causal Interactive Learning
AU - Nicholas Tagliapietra
AU - Florian Peter Busch
AU - Moritz Willig
AU - Matej Zečević
AU - Lavdim Halilaj
AU - Juergen Luettin
AU - Kristian Kersting
PY - 2026
DA - 2026/09/22
PB - arXiv
DO - 10.48550/arxiv.2609.26037
UR - https://arxiv.org/abs/2609.26037
AB - Explanations are central to causal reasoning, and cognitive science has long established that the human drive to explain is itself a mechanism for learning about causality. Despite this, learning from those abductive signals is largely ignored in artificial intelligence. While explainable AI (XAI) increasingly draws on causal models to generate explanations, the converse direction about what explanations can do for causality remains largely unexplored. To fill this gap, we propose xWhyL, a formal framework connecting causality and XAI by learning causal models from explanations. We develop a mathematical theory that translates explanations into a learning signal complementary to observational data, and demonstrate how it enables overcoming the limits of observational causal discovery. As explanations can be derived from incorrect beliefs and clash with data, a tension we call the Causal Tug-of-War, we prove conditions under which our framework rejects misspecified explanations rather than absorbing them. Our practical instantiation, Causal Interactive Learning (CIL), shows how expert explanations can efficiently support causal discovery and distinguish correct from incorrect explanations.
ER -