Semantic Loss Guided Data Efficient Supervised Fine Tuning for Safe Responses in LLMs

• URL: http://arxiv.org/abs/2412.06843v2
• Date: Wed, 11 Dec 2024 12:35:25 GMT
• Title: Semantic Loss Guided Data Efficient Supervised Fine Tuning for Safe Responses in LLMs
• Authors: Yuxiao Lu, Arunesh Sinha, Pradeep Varakantham,
• Abstract summary: Large Language Models (LLMs) generating unsafe responses to toxic prompts is a significant issue in their applications.<n>In this paper, we aim to take this problem and overcome limitations of requiring significant high-quality human data.<n>By employing a semantic cost combined with a negative Earth Mover Distance (EMD) loss, we guide the LLM away from generating unsafe responses.
• Score: 18.044879441434432
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Language Models (LLMs) generating unsafe responses to toxic prompts is a significant issue in their applications. While various efforts aim to address this safety concern, previous approaches often demand substantial human data collection or rely on the less dependable option of using another LLM to generate corrective data. In this paper, we aim to take this problem and overcome limitations of requiring significant high-quality human data. Our method requires only a small set of unsafe responses to toxic prompts, easily obtained from the unsafe LLM itself. By employing a semantic cost combined with a negative Earth Mover Distance (EMD) loss, we guide the LLM away from generating unsafe responses. Additionally, we propose a novel lower bound for EMD loss, enabling more efficient optimization. Our results demonstrate superior performance and data efficiency compared to baselines, and we further examine the nuanced effects of over-alignment and potential degradation of language capabilities when using contrastive data.

Related papers

• Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models? [83.53005932513155]
  Multi-modal large language models (MLLMs) have made significant progress, yet their safety alignment remains limited. We propose finetuning MLLMs on a small set of benign instruct-following data with responses replaced by simple, clear rejection sentences.
  arXiv  Detail & Related papers  (2025-04-14T09:03:51Z)
• Towards Harmless Multimodal Assistants with Blind Preference Optimization [49.044737689613164]
  Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities in multimodal understanding, reasoning, and interaction. Due to the effectiveness of preference optimization in aligning MLLMs with human preferences, there is an urgent need for safety-related preference data for MLLMs. We construct the MMSafe-PO preference dataset towards harmless multimodal assistants, featuring multimodal instructions, the conversational format, and ranked paired responses from human feedback.
  arXiv  Detail & Related papers  (2025-03-18T12:02:38Z)
• LLM-Safety Evaluations Lack Robustness [58.334290876531036]
  We argue that current safety alignment research efforts for large language models are hindered by many intertwined sources of noise. We propose a set of guidelines for reducing noise and bias in evaluations of future attack and defense papers.
  arXiv  Detail & Related papers  (2025-03-04T12:55:07Z)
• Refining Positive and Toxic Samples for Dual Safety Self-Alignment of LLMs with Minimal Human Interventions [17.485655062129965]
  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.
  arXiv  Detail & Related papers  (2025-02-08T09:54:47Z)
• Reward-Augmented Data Enhances Direct Preference Alignment of LLMs [56.24431208419858]
  We introduce reward-conditioned Large Language Models (LLMs) that learn from the entire spectrum of response quality within the dataset. We propose an effective yet simple data relabeling method that conditions the preference pairs on quality scores to construct a reward-augmented dataset.
  arXiv  Detail & Related papers  (2024-10-10T16:01:51Z)
• SLM as Guardian: Pioneering AI Safety with Small Language Models [6.799423428734095]
  Internalizing safeguard features into larger models brought challenges of higher training cost and unintended degradation of helpfulness. In this paper, we leverage a smaller LLM for both harmful query detection and safeguard response generation. We demonstrate the effectiveness of our approach, providing on par or surpassing harmful query detection and safeguard response performance compared to the publicly available LLMs.
  arXiv  Detail & Related papers  (2024-05-30T08:03:15Z)
• 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)
• Evaluation and Improvement of Fault Detection for Large Language Models [30.760472387136954]
  This paper investigates the effectiveness of existing fault detection methods for large language models (LLMs) We propose textbfMuCS, a prompt textbfMutation-based prediction textbfConfidence textbfSmoothing framework to boost the fault detection capability of existing methods.
  arXiv  Detail & Related papers  (2024-04-14T07:06:12Z)
• Illuminating Blind Spots of Language Models with Targeted Agent-in-the-Loop Synthetic Data [9.982616173090264]
  Language models (LMs) have achieved impressive accuracy across a variety of tasks but remain vulnerable to high-confidence misclassifications (UUs) UUs cluster into blind spots in the feature space, leading to significant risks in high-stakes applications. We propose a novel approach to address blind spot mitigation through the use of intelligent agents as teachers to characterize UU-type errors.
  arXiv  Detail & Related papers  (2024-03-26T16:49:25Z)
• 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)
• 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)
• On the Risk of Misinformation Pollution with Large Language Models [127.1107824751703]
  We investigate the potential misuse of modern Large Language Models (LLMs) for generating credible-sounding misinformation. Our study reveals that LLMs can act as effective misinformation generators, leading to a significant degradation in the performance of Open-Domain Question Answering (ODQA) systems.
  arXiv  Detail & Related papers  (2023-05-23T04:10:26Z)

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.