BARRIER: Bounded Activation Regions for Robust Information Erasure

TL;DR

BARRIER introduces a novel geometric framework for robust machine unlearning by bounding activation regions, enabling targeted concept erasure with formal guarantees and minimal collateral damage.

Contribution

It shifts the intervention focus from model weights to activation space geometry, employing Interval Arithmetic for rigorous and aggressive unlearning updates.

Findings

Matches state-of-the-art trade-offs in concept erasure

Provides formal bounds on model response during unlearning

Ensures preservation of neutral concepts with aggressive updates

Abstract

Machine unlearning has reached a critical bottleneck. As traditional weight-space interventions focus primarily on erasing targeted concepts, they often fail to prevent the unintended suppression of other significant representations. This leads to substantial collateral damage, with essential knowledge being forgotten, because these methods lack formal mathematical guarantees for the preservation of neutral concepts. To avoid degradation, they are frequently forced into conservative updates. We propose BARRIER (Bounded Activation Regions for Robust Information Erasure), a paradigm-shifting framework that shifts the locus of intervention from static model weights to the dynamic geometry of hidden-layer activations. Unlike existing methods, BARRIER employs Interval Arithmetic (IA) on SVD-based projections of the activation space to encapsulate the specific target region within a bounding hypercube. By driving unlearning updates exclusively within this forget interval and mathematically bounding the model response on the complement, we ensure rigorous protection of the retain distribution. This geometric construction transforms the preservation of knowledge from an empirical heuristic into a formal optimization target with a probabilistic tail bound on functional drift. Crucially, this stability permits highly aggressive unlearning updates within the forget region. Empirical evaluations demonstrate that BARRIER matches state-of-the-art trade-offs across classifiers and diffusion models, maximizing targeted concept erasure while safeguarding the integrity of all other representations. Our code is available at https://github.com/OneAndZero24/BARRIER.