Refining Positive and Toxic Samples for Dual Safety Self-Alignment of LLMs with Minimal Human Interventions

• URL: http://arxiv.org/abs/2502.08657v1
• Date: Sat, 08 Feb 2025 09:54:47 GMT
• Title: Refining Positive and Toxic Samples for Dual Safety Self-Alignment of LLMs with Minimal Human Interventions
• Authors: Jingxin Xu, Guoshun Nan, Sheng Guan, Sicong Leng, Yilian Liu, Zixiao Wang, Yuyang Ma, Zhili Zhou, Yanzhao Hou, Xiaofeng Tao,
• Abstract summary: Recent AI agents rely on instruction tuning and reinforcement learning to calibrate the output of large language models (LLMs) with human intentions. We propose PT-ALIGN, a novel safety self-alignment approach that minimizes human supervision by automatically refining positive and toxic samples. Experiments on 9 popular open-source LLMs demonstrate the effectiveness of our PT-ALIGN for safety alignment, while maintaining comparable levels of helpfulness and usefulness.
• Score: 17.485655062129965
• License:
• Abstract: Recent AI agents, such as ChatGPT and LLaMA, primarily rely on instruction tuning and reinforcement learning to calibrate the output of large language models (LLMs) with human intentions, ensuring the outputs are harmless and helpful. Existing methods heavily depend on the manual annotation of high-quality positive samples, while contending with issues such as noisy labels and minimal distinctions between preferred and dispreferred response data. However, readily available toxic samples with clear safety distinctions are often filtered out, removing valuable negative references that could aid LLMs in safety alignment. In response, we propose PT-ALIGN, a novel safety self-alignment approach that minimizes human supervision by automatically refining positive and toxic samples and performing fine-grained dual instruction tuning. Positive samples are harmless responses, while toxic samples deliberately contain extremely harmful content, serving as a new supervisory signals. Specifically, we utilize LLM itself to iteratively generate and refine training instances by only exploring fewer than 50 human annotations. We then employ two losses, i.e., maximum likelihood estimation (MLE) and fine-grained unlikelihood training (UT), to jointly learn to enhance the LLM's safety. The MLE loss encourages an LLM to maximize the generation of harmless content based on positive samples. Conversely, the fine-grained UT loss guides the LLM to minimize the output of harmful words based on negative samples at the token-level, thereby guiding the model to decouple safety from effectiveness, directing it toward safer fine-tuning objectives, and increasing the likelihood of generating helpful and reliable content. Experiments on 9 popular open-source LLMs demonstrate the effectiveness of our PT-ALIGN for safety alignment, while maintaining comparable levels of helpfulness and usefulness.

Related papers

• Look Before You Leap: Enhancing Attention and Vigilance Regarding Harmful Content with GuidelineLLM [53.79753074854936]
  Large language models (LLMs) are increasingly vulnerable to emerging jailbreak attacks. This vulnerability poses significant risks to the real-world applications. We propose a novel defensive paradigm called GuidelineLLM.
  arXiv  Detail & Related papers  (2024-12-10T12:42:33Z)
• Semantic Loss Guided Data Efficient Supervised Fine Tuning for Safe Responses in LLMs [18.044879441434432]
  Large Language Models (LLMs) generating unsafe responses to toxic prompts is a significant issue in their applications. In this paper, we aim to take this problem and overcome limitations of requiring significant high-quality human data. By employing a semantic cost combined with a negative Earth Mover Distance (EMD) loss, we guide the LLM away from generating unsafe responses.
  arXiv  Detail & Related papers  (2024-12-07T16:35:14Z)
• Large Language Models can be Strong Self-Detoxifiers [82.6594169242814]
  Self-disciplined Autoregressive Sampling (SASA) is a lightweight controlled decoding algorithm for toxicity reduction of large language models (LLMs) SASA tracks the margin of the current output to steer the generation away from the toxic subspace, by adjusting the autoregressive sampling strategy. evaluated on LLMs of different scale and nature, namely Llama-3.1-Instruct (8B), Llama-2 (7B), and GPT2-L models with the RealToxicityPrompts, BOLD, and AttaQ benchmarks.
  arXiv  Detail & Related papers  (2024-10-04T17:45:15Z)
• Bi-Factorial Preference Optimization: Balancing Safety-Helpfulness in Language Models [94.39278422567955]
  Fine-tuning large language models (LLMs) on human preferences has proven successful in enhancing their capabilities. However, ensuring the safety of LLMs during the fine-tuning remains a critical concern. We propose a supervised learning framework called Bi-Factorial Preference Optimization (BFPO) to address this issue.
  arXiv  Detail & Related papers  (2024-08-27T17:31:21Z)
• Robustifying Safety-Aligned Large Language Models through Clean Data Curation [11.273749179260468]
  Large language models (LLMs) are vulnerable when trained on datasets containing harmful content. In this paper, we propose a data curation framework designed to counter adversarial impacts in both scenarios.
  arXiv  Detail & Related papers  (2024-05-24T04:50:38Z)
• A Framework for Real-time Safeguarding the Text Generation of Large Language Model [12.683042228674694]
  Large Language Models (LLMs) have significantly advanced natural language processing (NLP) tasks. They pose ethical and societal risks due to their propensity to generate harmful content. We propose LLMSafeGuard, a lightweight framework to safeguard LLM text generation in real-time.
  arXiv  Detail & Related papers  (2024-04-29T18:40:01Z)
• Negating Negatives: Alignment with Human Negative Samples via Distributional Dispreference Optimization [37.8788435790632]
  Large language models (LLMs) have revolutionized the role of AI, yet pose potential social risks. Existing methods rely on high-quality positive-negative training pairs, suffering from noisy positive responses that are barely distinguishable from negative ones. We propose Distributional Dispreference Optimization (D$2$O), which maximizes the discrepancy between dispreferred responses and the generated non-negative ones.
  arXiv  Detail & Related papers  (2024-03-06T03:02:38Z)
• Making Harmful Behaviors Unlearnable for Large Language Models [50.44915524846857]
  Large language models (LLMs) have shown great potential as general-purpose AI assistants in various domains. LLMs can be easily fine-tuned into harmful assistants as the fine-tuning data often contains implicit or explicit harmful content. This paper proposes a controllable training framework that makes harmful behaviors unlearnable during the fine-tuning process.
  arXiv  Detail & Related papers  (2023-11-02T09:18:21Z)
• Gaining Wisdom from Setbacks: Aligning Large Language Models via Mistake Analysis [127.85293480405082]
  The rapid development of large language models (LLMs) has not only provided numerous opportunities but also presented significant challenges. Existing alignment methods usually direct LLMs toward the favorable outcomes by utilizing human-annotated, flawless instruction-response pairs. This study proposes a novel alignment technique based on mistake analysis, which deliberately exposes LLMs to erroneous content to learn the reasons for mistakes and how to avoid them.
  arXiv  Detail & Related papers  (2023-10-16T14:59:10Z)
• Do-Not-Answer: A Dataset for Evaluating Safeguards in LLMs [59.596335292426105]
  This paper collects the first open-source dataset to evaluate safeguards in large language models. We train several BERT-like classifiers to achieve results comparable with GPT-4 on automatic safety evaluation.
  arXiv  Detail & Related papers  (2023-08-25T14:02:12Z)

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.