# Bias Redistribution in Visual Machine Unlearning: Does Forgetting One Group Harm Another?

arXiv: https://arxiv.org/abs/2604.08111 · HTML: https://mlsanity.com/paper/2604.08111/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Machine unlearning, one of 12 directions in LLM safety and security.
- Builds on 23 of its 31 references that are in the map.
- First posted on arXiv 9 Apr 2026.

## Placed in

- LLM safety and security / [Machine unlearning](https://mlsanity.com/safety/machine-unlearning/)

## Builds on (most cited first, 23 of 23)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Deep Learning Face Attributes in the Wild](https://mlsanity.com/paper/1411.7766/) (2014)
- [Machine Unlearning](https://mlsanity.com/paper/1912.03817/) (2019)
- [Refusal in Language Models Is Mediated by a Single Direction](https://mlsanity.com/paper/2406.11717/) (2024)
- [Erasing Concepts from Diffusion Models](https://mlsanity.com/paper/2303.07345/) (2023)
- [Null It Out: Guarding Protected Attributes by Iterative Nullspace Projection](https://mlsanity.com/paper/2004.07667/) (2020)
- [SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation](https://mlsanity.com/paper/2310.12508/) (2023)
- [Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness](https://mlsanity.com/paper/1711.05144/) (2017)
- [Towards Unbounded Machine Unlearning](https://mlsanity.com/paper/2302.09880/) (2023)
- [LEACE: Perfect linear concept erasure in closed form](https://mlsanity.com/paper/2306.03819/) (2023)
- [Machine Unlearning of Features and Labels](https://mlsanity.com/paper/2108.11577/) (2021)
- [Fast Machine Unlearning Without Retraining Through Selective Synaptic Dampening](https://mlsanity.com/paper/2308.07707/) (2023)
- [DeAR: Debiasing Vision-Language Models with Additive Residuals](https://mlsanity.com/paper/2303.10431/) (2023)
- [FairCLIP: Harnessing Fairness in Vision-Language Learning](https://mlsanity.com/paper/2403.19949/) (2024)
- [A Survey on Federated Unlearning: Challenges, Methods, and Future Directions](https://mlsanity.com/paper/2310.20448/) (2023)
- [FACET: Fairness in Computer Vision Evaluation Benchmark](https://mlsanity.com/paper/2309.00035/) (2023)
- [Spiking Vision Transformer with Saccadic Attention](https://mlsanity.com/paper/2502.12677/) (2025)
- [FairCLIP: Social Bias Elimination based on Attribute Prototype Learning and Representation Neutralization](https://mlsanity.com/paper/2210.14562/) (2022)
- [Learning Global Hypothesis Space for Enhancing Synergistic Reasoning Chain](https://mlsanity.com/paper/2602.09794/) (2026)
- [Deep Unlearn: Benchmarking Machine Unlearning for Image Classification](https://mlsanity.com/paper/2410.01276/) (2024)
- [DatasetAgent: A Novel Multi-Agent System for Auto-Constructing Datasets from Real-World Images](https://mlsanity.com/paper/2507.08648/) (2025)
- [Group-robust Machine Unlearning](https://mlsanity.com/paper/2503.09330/) (2025)
- [SD-PSFNet: Sequential and Dynamic Point Spread Function Network for Image Deraining](https://mlsanity.com/paper/2511.17993/) (2025)

## Abstract (arXiv)

Machine unlearning enables models to selectively forget training data, driven by privacy regulations such as GDPR and CCPA. However, its fairness implications remain underexplored: when a model forgets a demographic group, does it neutralize that concept or redistribute it to correlated groups, potentially amplifying bias? We investigate this bias redistribution phenomenon on CelebA using CLIP models (ViT/B-32, ViT-L/14, ViT-B/16) under a zero-shot classification setting across intersectional groups defined by age and gender. We evaluate three unlearning methods, Prompt Erasure, Prompt Reweighting, and Refusal Vector using per-group accuracy shifts, demographic parity gaps, and a redistribution score. Our results show that unlearning does not eliminate bias but redistributes it primarily along gender rather than age boundaries. In particular, removing the dominant Young Female group consistently transfers performance to Old Female across all model scales, revealing a gender-dominant structure in CLIP's embedding space. While the Refusal Vector method reduces redistribution, it fails to achieve complete forgetting and significantly degrades retained performance. These findings highlight a fundamental limitation of current unlearning methods: without accounting for embedding geometry, they risk amplifying bias in retained groups.