SOURCE-LINKED INTELLIGENCE

MUtE: A Dual Framework for Concept Erasure and Counterfactual Interventions

arXiv · AI, language, vision and robotics · article · Sep 10, 2026 · UTC

Erasing concept-specific information from representations has been proven useful for mitigating bias or interpreting model decisions. The joint objective is to transform the original representations such that the target concept becomes unpredictable, while maximally preserving concept-unrelated information. In this work, we revisit the optimal bounds of concept erasure to derive a novel class of erasure functions that naturally induce a deterministic, dual counterfactual mapping. Bridging the gap between theoretical optimality and practical representation learning, we design an implementation

Read original source ↗ Open in workspace

recordType

paper

region

Global