Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models?

• URL: http://arxiv.org/abs/2504.10000v1
• Date: Mon, 14 Apr 2025 09:03:51 GMT
• Title: Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models?
• Authors: Yanbo Wang, Jiyang Guan, Jian Liang, Ran He,
• Abstract summary: Multi-modal large language models (MLLMs) have made significant progress, yet their safety alignment remains limited.<n>We propose finetuning MLLMs on a small set of benign instruct-following data with responses replaced by simple, clear rejection sentences.
• Score: 83.53005932513155
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Multi-modal large language models (MLLMs) have made significant progress, yet their safety alignment remains limited. Typically, current open-source MLLMs rely on the alignment inherited from their language module to avoid harmful generations. However, the lack of safety measures specifically designed for multi-modal inputs creates an alignment gap, leaving MLLMs vulnerable to vision-domain attacks such as typographic manipulation. Current methods utilize a carefully designed safety dataset to enhance model defense capability, while the specific knowledge or patterns acquired from the high-quality dataset remain unclear. Through comparison experiments, we find that the alignment gap primarily arises from data distribution biases, while image content, response quality, or the contrastive behavior of the dataset makes little contribution to boosting multi-modal safety. To further investigate this and identify the key factors in improving MLLM safety, we propose finetuning MLLMs on a small set of benign instruct-following data with responses replaced by simple, clear rejection sentences. Experiments show that, without the need for labor-intensive collection of high-quality malicious data, model safety can still be significantly improved, as long as a specific fraction of rejection data exists in the finetuning set, indicating the security alignment is not lost but rather obscured during multi-modal pretraining or instruction finetuning. Simply correcting the underlying data bias could narrow the safety gap in the vision domain.

Related papers

• Fundamental Safety-Capability Trade-offs in Fine-tuning Large Language Models [92.38300626647342]
  Fine-tuning Large Language Models (LLMs) on some task-specific datasets has been a primary use of LLMs. This paper presents a theoretical framework for understanding the interplay between safety and capability in two primary safety-aware LLM fine-tuning strategies.
  arXiv  Detail & Related papers  (2025-03-24T20:41:57Z)
• MLLM-as-a-Judge for Image Safety without Human Labeling [81.24707039432292]
  In the age of AI-generated content (AIGC), many image generation models are capable of producing harmful content. It is crucial to identify such unsafe images based on established safety rules. Existing approaches typically fine-tune MLLMs with human-labeled datasets.
  arXiv  Detail & Related papers  (2024-12-31T00:06:04Z)
• Multitask Mayhem: Unveiling and Mitigating Safety Gaps in LLMs Fine-tuning [1.3307486544794784]
  Red teaming/Safety alignment efforts show that fine-tuning models on benign (non-harmful) data could compromise safety. This paper explores the task-wise safety degradation due to fine-tuning on downstream tasks such as summarization, code generation, translation, and classification. Our work underscores the need for generalized alignment measures to ensure safer and more robust models.
  arXiv  Detail & Related papers  (2024-09-18T08:04:24Z)
• CoCA: Regaining Safety-awareness of Multimodal Large Language Models with Constitutional Calibration [90.36429361299807]
  multimodal large language models (MLLMs) have demonstrated remarkable success in engaging in conversations involving visual inputs. The integration of visual modality has introduced a unique vulnerability: the MLLM becomes susceptible to malicious visual inputs. We introduce a technique termed CoCA, which amplifies the safety-awareness of the MLLM by calibrating its output distribution.
  arXiv  Detail & Related papers  (2024-09-17T17:14:41Z)
• Cross-modality Information Check for Detecting Jailbreaking in Multimodal Large Language Models [17.663550432103534]
  Multimodal Large Language Models (MLLMs) extend the capacity of LLMs to understand multimodal information comprehensively. These models are susceptible to jailbreak attacks, where malicious users can break the safety alignment of the target model and generate misleading and harmful answers. We propose Cross-modality Information DEtectoR (CIDER), a plug-and-play jailbreaking detector designed to identify maliciously perturbed image inputs.
  arXiv  Detail & Related papers  (2024-07-31T15:02:46Z)
• SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model [77.86593720792986]
  We propose a Safety Preference Alignment dataset for Vision Language Models named SPA-VL.<n> SPA-VL covers 6 harmfulness domains, 13 categories, and 53 subcategories, and contains 100,788 samples of the quadruple (question, image, chosen response, rejected response)<n>Experiments indicate that models trained with alignment techniques on the SPA-VL dataset exhibit substantial improvements in harmlessness and helpfulness while maintaining core capabilities.
  arXiv  Detail & Related papers  (2024-06-17T18:57:37Z)
• Fine-Tuning, Quantization, and LLMs: Navigating Unintended Outcomes [0.0]
  Large Language Models (LLMs) have gained widespread adoption across various domains, including chatbots and auto-task completion agents. These models are susceptible to safety vulnerabilities such as jailbreaking, prompt injection, and privacy leakage attacks. This study investigates the impact of these modifications on LLM safety, a critical consideration for building reliable and secure AI systems.
  arXiv  Detail & Related papers  (2024-04-05T20:31:45Z)
• RigorLLM: Resilient Guardrails for Large Language Models against Undesired Content [62.685566387625975]
  Current mitigation strategies, while effective, are not resilient under adversarial attacks. This paper introduces Resilient Guardrails for Large Language Models (RigorLLM), a novel framework designed to efficiently moderate harmful and unsafe inputs.
  arXiv  Detail & Related papers  (2024-03-19T07:25:02Z)
• Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models [39.56233272612982]
  Current vision large language models (VLLMs) exhibit remarkable capabilities yet are prone to generate harmful content and are vulnerable to jailbreaking attacks. Our initial analysis finds that this is due to the presence of harmful data during vision-language instruction fine-tuning. To address this issue, we first curate a vision-language safe instruction-following dataset VLGuard covering various harmful categories.
  arXiv  Detail & Related papers  (2024-02-03T16:43:42Z)

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.