Modality-Aware Neuron Pruning for Unlearning in Multimodal Large Language Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908461622624256 |
|---|---|
| author | Liu, Zheyuan Dou, Guangyao Yuan, Xiangchi Zhang, Chunhui Tan, Zhaoxuan Jiang, Meng |
| author_facet | Liu, Zheyuan Dou, Guangyao Yuan, Xiangchi Zhang, Chunhui Tan, Zhaoxuan Jiang, Meng |
| contents | Generative models such as Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) trained on massive datasets can lead them to memorize and inadvertently reveal sensitive information, raising ethical and privacy concerns. While some prior works have explored this issue in the context of LLMs, it presents a unique challenge for MLLMs due to the entangled nature of knowledge across modalities, making comprehensive unlearning more difficult. To address this challenge, we propose Modality Aware Neuron Unlearning (MANU), a novel unlearning framework for MLLMs designed to selectively clip neurons based on their relative importance to the targeted forget data, curated for different modalities. Specifically, MANU consists of two stages: important neuron selection and selective pruning. The first stage identifies and collects the most influential neurons across modalities relative to the targeted forget knowledge, while the second stage is dedicated to pruning those selected neurons. MANU effectively isolates and removes the neurons that contribute most to the forget data within each modality, while preserving the integrity of retained knowledge. Our experiments conducted across various MLLM architectures illustrate that MANU can achieve a more balanced and comprehensive unlearning in each modality without largely affecting the overall model utility. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_15910 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Modality-Aware Neuron Pruning for Unlearning in Multimodal Large Language Models Liu, Zheyuan Dou, Guangyao Yuan, Xiangchi Zhang, Chunhui Tan, Zhaoxuan Jiang, Meng Computation and Language Generative models such as Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) trained on massive datasets can lead them to memorize and inadvertently reveal sensitive information, raising ethical and privacy concerns. While some prior works have explored this issue in the context of LLMs, it presents a unique challenge for MLLMs due to the entangled nature of knowledge across modalities, making comprehensive unlearning more difficult. To address this challenge, we propose Modality Aware Neuron Unlearning (MANU), a novel unlearning framework for MLLMs designed to selectively clip neurons based on their relative importance to the targeted forget data, curated for different modalities. Specifically, MANU consists of two stages: important neuron selection and selective pruning. The first stage identifies and collects the most influential neurons across modalities relative to the targeted forget knowledge, while the second stage is dedicated to pruning those selected neurons. MANU effectively isolates and removes the neurons that contribute most to the forget data within each modality, while preserving the integrity of retained knowledge. Our experiments conducted across various MLLM architectures illustrate that MANU can achieve a more balanced and comprehensive unlearning in each modality without largely affecting the overall model utility. |
| title | Modality-Aware Neuron Pruning for Unlearning in Multimodal Large Language Models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2502.15910 |