Exploring and Bridging Knowledge Holes in Unlearned Multimodal Large Language Models

• URL: http://arxiv.org/abs/2608.01849v1
• Date: Mon, 03 Aug 2026 07:57:34 GMT
• Title: Exploring and Bridging Knowledge Holes in Unlearned Multimodal Large Language Models
• Abstract summary: We propose a benchmark that captures unintended degradation on benign inputs sharing generic patterns with the forget set.<n>We also propose Selective Protection with Anchored Regularization, which protects generic patterns via anchored activation filtering.<n>Our experiments on SafeEraser demonstrate that SPAR recovers over 98% of vanilla response quality compared to below 50% for standard baselines.
• Score: 23.51933427959403
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine unlearning offers a promising approach to remove unsafe content from Multimodal Large Language Models (MLLMs), yet ensuring the precision of unlearning remains a persistent challenge. One reason is that current MLLM unlearning evaluation paradigms suffer from a critical blind spot: they assess model utility through benchmarks whose representations are distant from the forget set, failing to capture knowledge holes---severe degradation on benign adjacent inputs. To probe knowledge holes in unlearned MLLMs, we construct a benchmark that captures unintended degradation on benign inputs sharing generic patterns with the forget set, and confirm through controlled experiments that they are a systematic consequence of commonly used approaches. Furthermore, to bridge this gap, we propose Selective Protection with Anchored Regularization, which protects generic patterns via anchored activation filtering while reinforcing them through entity-abstracted enhancement. Our experiments on SafeEraser demonstrate that SPAR recovers over 98% of vanilla response quality compared to below 50% for standard baselines---while achieving 0.00% attack success rate and competitive model utility. These results underscore the necessity of more fine-grained evaluation for trustworthy MLLM unlearning.

Related papers

• Test-Time Unlearning via Sparse Autoencoder [55.4290003355764]
  We propose ARIA, a test-time unlearning method that leaves model weights intact and gates access to unwanted knowledge only when generation enters a forget-related state.<n>ARIA improves the forget-retain trade-off over weight-based baselines across both a thinking model and an instruction model.
  arXiv  Detail & Related papers  (2026-09-14T18:58:09Z)
• Measuring the Depth of LLM Unlearning via Activation Patching [3.590648050085134]
  Large language model (LLM) unlearning has emerged as a crucial post-hoc mechanism for privacy protection and AI safety.<n>Existing output-level metrics fail to detect when this knowledge remains recoverable from internal representations.<n>We propose the Unlearning Depth Score (UDS), a metric that quantifies the mechanistic depth of unlearning via activation patching.
  arXiv  Detail & Related papers  (2026-05-23T14:52:28Z)
• Leak@$k$: Unlearning Does Not Make LLMs Forget Under Probabilistic Decoding [18.830386174815583]
  We show that textitalmost all existing unlearning methods fail to achieve true forgetting in practice.<n>We introduce textttleak@$k$, a new meta-evaluation metric that quantifies the likelihood of forgotten knowledge reappearing.
  arXiv  Detail & Related papers  (2025-11-07T02:30:05Z)
• Unlearning That Lasts: Utility-Preserving, Robust, and Almost Irreversible Forgetting in LLMs [31.768387661474904]
  Unlearning in large language models (LLMs) involves precisely removing specific information from a pre-trained model.<n>This is crucial to ensure safety of LLMs by deleting private data or harmful knowledge acquired during pre-training.<n>We introduce JensUn, where we leverage the Jensen-Shannon Divergence as the training objective for both forget and retain sets.<n>In extensive experiments, JensUn achieves better forget-utility trade-off than competing methods, and even demonstrates strong resilience to benign relearning.
  arXiv  Detail & Related papers  (2025-09-02T20:38:53Z)
• One Token to Fool LLM-as-a-Judge [52.45386385722788]
  Large language models (LLMs) are increasingly trusted as automated judges, assisting evaluation and providing reward signals for training other models.<n>We uncover a critical vulnerability even in this reference-based paradigm: generative reward models are systematically susceptible to reward hacking.
  arXiv  Detail & Related papers  (2025-07-11T17:55:22Z)
• SafeEraser: Enhancing Safety in Multimodal Large Language Models through Multimodal Machine Unlearning [48.42431979599426]
  We propose SAFEERASER, a safety unlearning benchmark for Multimodal Large Language Models (MLLMs)<n>We comprehensively evaluate unlearning methods from two perspectives: forget quality and model utility.<n> Experimental results demonstrate that combining PD Loss with existing unlearning methods can effectively prevent over-forgetting.
  arXiv  Detail & Related papers  (2025-02-18T04:09:46Z)
• Unconditional Truthfulness: Learning Unconditional Uncertainty of Large Language Models [104.55763564037831]
  We train a regression model that leverages attention maps, probabilities on the current generation step, and recurrently computed uncertainty scores from previously generated tokens.<n>Our evaluation shows that the proposed method is highly effective for selective generation, achieving substantial improvements over rivaling unsupervised and supervised approaches.
  arXiv  Detail & Related papers  (2024-08-20T09:42:26Z)
• Towards Effective Evaluations and Comparisons for LLM Unlearning Methods [97.2995389188179]
  This paper seeks to refine the evaluation of machine unlearning for large language models.<n>It addresses two key challenges -- the robustness of evaluation metrics and the trade-offs between competing goals.
  arXiv  Detail & Related papers  (2024-06-13T14:41:00Z)
• Large Language Model Unlearning via Embedding-Corrupted Prompts [10.889859281637406]
  We present textbfEmbedding-COrrupted (ECO) Prompts, a lightweight unlearning framework for large language models. We enforce an unlearned state during inference by employing a prompt classifier to identify and safeguard prompts to forget. We find that these embedding-corrupted prompts not only lead to desirable outputs that satisfy the unlearning objective but also closely approximate the output from a model that has never been trained on the data intended for forgetting.
  arXiv  Detail & Related papers  (2024-06-12T06:56:20Z)
• ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks [91.55895047448249]
  This paper presents ReEval, an LLM-based framework using prompt chaining to perturb the original evidence for generating new test cases. We implement ReEval using ChatGPT and evaluate the resulting variants of two popular open-domain QA datasets. Our generated data is human-readable and useful to trigger hallucination in large language models.
  arXiv  Detail & Related papers  (2023-10-19T06:37:32Z)

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.