EvoMU: Evolutionary Machine Unlearning

• URL: http://arxiv.org/abs/2602.02139v2
• Date: Mon, 09 Feb 2026 10:08:08 GMT
• Title: EvoMU: Evolutionary Machine Unlearning
• Authors: Pawel Batorski, Paul Swoboda,
• Abstract summary: EvoMU finds task-specific losses in the vast space of possible unlearning loss functions.<n>This work is therefore an instance of automatic scientific discovery, a.k.a. an AI co-scientist.
• Score: 13.775690509818753
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine unlearning aims to unlearn specified training data (e.g. sensitive or copyrighted material). A prominent approach is to fine-tune an existing model with an unlearning loss that retains overall utility. The space of suitable unlearning loss functions is vast, making the search for an optimal loss function daunting. Additionally, there might not even exist a universally optimal loss function: differences in the structure and overlap of the forget and retain data can cause a loss to work well in one setting but over-unlearn or under-unlearn in another. Our approach EvoMU tackles these two challenges simultaneously. An evolutionary search procedure automatically finds task-specific losses in the vast space of possible unlearning loss functions. This allows us to find dataset-specific losses that match or outperform existing losses from the literature, without the need for a human-in-the-loop. This work is therefore an instance of automatic scientific discovery, a.k.a. an AI co-scientist. In contrast to previous AI co-scientist works, we do so on a budget: We achieve SotA results using a small 4B parameter model (Qwen3-4B-Thinking), showing the potential of AI co-scientists with limited computational resources. Our experimental evaluation shows that we surpass previous loss-based unlearning formulations on TOFU-5%, TOFU-10%, MUSE and WMDP by synthesizing novel unlearning losses. Our code is available at https://github.com/Batorskq/EvoMU.

Related papers

• BLUR: A Bi-Level Optimization Approach for LLM Unlearning [100.90394814817965]
  We argue that it is important to model the hierarchical structure of the unlearning problem.<n>We propose a novel algorithm, termed Bi-Level UnleaRning (textttBLUR), which delivers superior performance.
  arXiv  Detail & Related papers  (2025-06-09T19:23:05Z)
• What should an AI assessor optimise for? [57.96463917842822]
  An AI assessor is an external, ideally indepen-dent system that predicts an indicator, e.g., a loss value, of another AI system.<n>Here we address the question: is it always optimal to train the assessor for the target metric?<n>We experimentally explore this question for, respectively, regression losses and classification scores with monotonic and non-monotonic mappings.
  arXiv  Detail & Related papers  (2025-02-01T08:41:57Z)
• Attribute-to-Delete: Machine Unlearning via Datamodel Matching [65.13151619119782]
  Machine unlearning -- efficiently removing a small "forget set" training data on a pre-divertrained machine learning model -- has recently attracted interest. Recent research shows that machine unlearning techniques do not hold up in such a challenging setting.
  arXiv  Detail & Related papers  (2024-10-30T17:20:10Z)
• On Using Admissible Bounds for Learning Forward Search Heuristics [9.749638953163391]
  We focus on how to effectively utilize the information provided by admissibles in learning. We model the learned as a truncated ssian, where admissibles are used not as training targets but as lower bounds of this distribution. Results show that our proposed method converges faster during training and yields bettergauss.
  arXiv  Detail & Related papers  (2023-08-23T04:14:45Z)
• Alternate Loss Functions for Classification and Robust Regression Can Improve the Accuracy of Artificial Neural Networks [6.452225158891343]
  This paper shows that training speed and final accuracy of neural networks can significantly depend on the loss function used to train neural networks. Two new classification loss functions that significantly improve performance on a wide variety of benchmark tasks are proposed.
  arXiv  Detail & Related papers  (2023-03-17T12:52:06Z)
• Xtreme Margin: A Tunable Loss Function for Binary Classification Problems [0.0]
  We provide an overview of a novel loss function, the Xtreme Margin loss function. Unlike the binary cross-entropy and the hinge loss functions, this loss function provides researchers and practitioners flexibility with their training process.
  arXiv  Detail & Related papers  (2022-10-31T22:39:32Z)
• Omnipredictors [19.735769148626588]
  Loss minimization is a dominant paradigm in machine learning. We introduce the notion of an ($mathcalL,mathcalC$)-omnipredictor, which could be used to optimize any loss in a family. We show that such "loss-oblivious'' learning is feasible through a connection to multicalibration.
  arXiv  Detail & Related papers  (2021-09-11T23:28:49Z)
• Loss Function Discovery for Object Detection via Convergence-Simulation Driven Search [101.73248560009124]
  We propose an effective convergence-simulation driven evolutionary search algorithm, CSE-Autoloss, for speeding up the search progress. We conduct extensive evaluations of loss function search on popular detectors and validate the good generalization capability of searched losses. Our experiments show that the best-discovered loss function combinations outperform default combinations by 1.1% and 0.8% in terms of mAP for two-stage and one-stage detectors.
  arXiv  Detail & Related papers  (2021-02-09T08:34:52Z)
• Learning by Minimizing the Sum of Ranked Range [58.24935359348289]
  We introduce the sum of ranked range (SoRR) as a general approach to form learning objectives. A ranked range is a consecutive sequence of sorted values of a set of real numbers. We explore two applications in machine learning of the minimization of the SoRR framework, namely the AoRR aggregate loss for binary classification and the TKML individual loss for multi-label/multi-class classification.
  arXiv  Detail & Related papers  (2020-10-05T01:58:32Z)
• Predicting Training Time Without Training [120.92623395389255]
  We tackle the problem of predicting the number of optimization steps that a pre-trained deep network needs to converge to a given value of the loss function. We leverage the fact that the training dynamics of a deep network during fine-tuning are well approximated by those of a linearized model. We are able to predict the time it takes to fine-tune a model to a given loss without having to perform any training.
  arXiv  Detail & Related papers  (2020-08-28T04:29:54Z)
• A Loss-Function for Causal Machine-Learning [0.0]
  Causal machine-learning is about predicting the net-effect (true-lift) of treatments. There is no similarly well-defined loss function due to the lack of point-wise true values in the data. We propose a novel method to define a loss function in this context, which is equal to mean-square-error (MSE) in a standard regression problem.
  arXiv  Detail & Related papers  (2020-01-02T21:22:18Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.