Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable

• URL: http://arxiv.org/abs/2503.00555v1
• Date: Sat, 01 Mar 2025 16:42:01 GMT
• Title: Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable
• Authors: Tiansheng Huang, Sihao Hu, Fatih Ilhan, Selim Furkan Tekin, Zachary Yahn, Yichang Xu, Ling Liu,
• Abstract summary: Safety alignment is an important procedure before the official deployment of a Large Language Model.<n>We show that there exists a trade-off between reasoning and safety capability with the sequential LRM production pipeline.<n>As a by-product, we curate a dataset called DirectRefusal, which might serve as an alternative dataset for safety alignment.
• Score: 7.140765245328677
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Safety alignment is an important procedure before the official deployment of a Large Language Model (LLM). While safety alignment has been extensively studied for LLM, there is still a large research gap for Large Reasoning Models (LRMs) that equip with improved reasoning capability. We in this paper systematically examine a simplified pipeline for producing safety aligned LRMs. With our evaluation of various LRMs, we deliver two main findings: i) Safety alignment can be done upon the LRM to restore its safety capability. ii) Safety alignment leads to a degradation of the reasoning capability of LRMs. The two findings show that there exists a trade-off between reasoning and safety capability with the sequential LRM production pipeline. The discovered trade-off, which we name Safety Tax, should shed light on future endeavors of safety research on LRMs. As a by-product, we curate a dataset called DirectRefusal, which might serve as an alternative dataset for safety alignment. Our source code is available at https://github.com/git-disl/Safety-Tax.

Related papers

• SafeMLRM: Demystifying Safety in Multi-modal Large Reasoning Models [50.34706204154244]
  Acquiring reasoning capabilities catastrophically degrades inherited safety alignment. Certain scenarios suffer 25 times higher attack rates. Despite tight reasoning-answer safety coupling, MLRMs demonstrate nascent self-correction.
  arXiv  Detail & Related papers  (2025-04-09T06:53:23Z)
• SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities [21.317245896641136]
  Long chain-of-thought (CoT) reasoning generates structured intermediate steps, enhancing reasoning capabilities.<n>Current research on large language model (LLM) safety usually focuses on short-answer responses, overlooking the long CoT style outputs of LRMs.
  arXiv  Detail & Related papers  (2025-02-17T16:57:56Z)
• STAIR: Improving Safety Alignment with Introspective Reasoning [44.780098674618614]
  We propose STAIR, a framework that integrates SafeTy Alignment with Itrospective Reasoning.<n>We show that STAIR effectively mitigates harmful outputs while better preserving helpfulness, compared to instinctive alignment strategies.<n>With test-time scaling, STAIR achieves a safety performance comparable to Claude-3.5 against popular jailbreak attacks.
  arXiv  Detail & Related papers  (2025-02-04T15:02:55Z)
• Internal Activation as the Polar Star for Steering Unsafe LLM Behavior [50.463399903987245]
  We introduce SafeSwitch, a framework that dynamically regulates unsafe outputs by monitoring and utilizing the model's internal states.<n>Our empirical results show that SafeSwitch reduces harmful outputs by over 80% on safety benchmarks while maintaining strong utility.
  arXiv  Detail & Related papers  (2025-02-03T04:23:33Z)
• SaLoRA: Safety-Alignment Preserved Low-Rank Adaptation [41.91948079316541]
  Recent studies have raised concerns that LoRA fine-tuning could potentially compromise the safety alignment in large language models.<n>In this paper, we propose Safety-alignment preserved Low-Rank Adaptation (SaLoRA)<n>Unlike previous LoRA methods and their variants, SaLoRA enables targeted modifications to LLMs without disrupting their original alignments.<n>Our experiments show that SaLoRA outperforms various adapters-based approaches across various evaluation metrics in different fine-tuning tasks.
  arXiv  Detail & Related papers  (2025-01-03T11:34:28Z)
• Separate the Wheat from the Chaff: A Post-Hoc Approach to Safety Re-Alignment for Fine-Tuned Language Models [30.93821289892195]
  We propose IRR (Identify, Remove, and Recalibrate for Safety Realignment) that performs safety realignment for LLMs.<n>The core of IRR is to identify and remove unsafe delta parameters from the fine-tuned models, while recalibrating the retained ones.<n>Our results demonstrate that IRR significantly enhances the safety performance of fine-tuned models on safety benchmarks, such as harmful queries and jailbreak attacks.
  arXiv  Detail & Related papers  (2024-12-15T03:58:38Z)
• Superficial Safety Alignment Hypothesis [8.297367440457508]
  We propose the Superficial Safety Alignment Hypothesis (SSAH), which posits that safety alignment should teach an otherwise unsafe model to choose the correct reasoning direction. We identify four types of attribute-critical components in safety-aligned large language models (LLMs) Our findings show that freezing certain safety-critical components 7.5% during fine-tuning allows the model to retain its safety attributes while adapting to new tasks.
  arXiv  Detail & Related papers  (2024-10-07T19:53:35Z)
• SCANS: Mitigating the Exaggerated Safety for LLMs via Safety-Conscious Activation Steering [56.92068213969036]
  Safety alignment is indispensable for Large Language Models (LLMs) to defend threats from malicious instructions.<n>Recent researches reveal safety-aligned LLMs prone to reject benign queries due to the exaggerated safety issue.<n>We propose a Safety-Conscious Activation Steering (SCANS) method to mitigate the exaggerated safety concerns.
  arXiv  Detail & Related papers  (2024-08-21T10:01:34Z)
• Refuse Whenever You Feel Unsafe: Improving Safety in LLMs via Decoupled Refusal Training [67.30423823744506]
  This study addresses a critical gap in safety tuning practices for Large Language Models (LLMs) We introduce a novel approach, Decoupled Refusal Training (DeRTa), designed to empower LLMs to refuse compliance to harmful prompts at any response position. DeRTa incorporates two novel components: (1) Maximum Likelihood Estimation with Harmful Response Prefix, which trains models to recognize and avoid unsafe content by appending a segment of harmful response to the beginning of a safe response, and (2) Reinforced Transition Optimization (RTO), which equips models with the ability to transition from potential harm to safety refusal consistently throughout the harmful
  arXiv  Detail & Related papers  (2024-07-12T09:36:33Z)
• Towards Comprehensive Post Safety Alignment of Large Language Models via Safety Patching [74.62818936088065]
  textscSafePatching is a novel framework for comprehensive PSA.<n>textscSafePatching achieves a more comprehensive PSA than baseline methods.<n>textscSafePatching demonstrates its superiority in continual PSA scenarios.
  arXiv  Detail & Related papers  (2024-05-22T16:51:07Z)
• Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To! [88.90694413503614]
  We find that the safety alignment of LLMs can be compromised by fine-tuning. We jailbreak GPT-3.5 Turbo's safety guardrails by fine-tuning it on only 10 such examples. We advocate for further research efforts toward reinforcing safety protocols for the custom fine-tuning of aligned LLMs.
  arXiv  Detail & Related papers  (2023-10-05T17:12:17Z)

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.