Research map / Probabilistic and scientific modeling / Causal discovery

xWhy: Causal Learning from Explanations

NeurIPS 2026 · Probabilistic and scientific modeling › Causal discovery · approach: Causal representation identifiability

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes xWhy, a framework that learns causal structure from explanations (abductive signals) rather than ignoring them, based only on the abstract provided.

Problem

Learning causal knowledge from abductive explanations is largely ignored in AI, while XAI mainly uses causal models to generate explanations rather than the reverse.

Approach

Treats the human drive to explain as a learning mechanism: uses explanations as a signal for causal learning, the converse of the usual causal-model-for-XAI direction.

Closest papers by meaning

• A New Approach to Backtracking Counterfactual Explanations: A Unified Causal Framework for Efficient Model Interpretability (ICML 2025 · similarity 0.87)
• Unveiling Causal Reasoning in Large Language Models: Reality or Mirage? (NeurIPS 2024 · similarity 0.87)
• Causality can systematically address the monsters under the bench(marks) (NeurIPS 2026 · similarity 0.86)
• Causal learning with the invariance principle (NeurIPS 2026 · similarity 0.86)
• CausalTab: Pretraining Across Causal Environments for Tabular Causal Discovery (NeurIPS 2026 · similarity 0.86)
• The third pillar of causal analysis? A measurement perspective on causal representations (NeurIPS 2025 · similarity 0.86)