Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing

• URL: http://arxiv.org/abs/2606.00686v1
• Date: Sat, 30 May 2026 11:49:42 GMT
• Title: Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing
• Abstract summary: This work challenges the orthodoxy that unsafe data must be discarded.<n>We introduce SafeMoE, a framework that isolates unsafe knowledge into domain-specific LowRank Adapters (LoRA experts) trained exclusively on harmful corpora.<n>Our findings suggest a paradigm in alignment: true safety requires not the masking of unsafe knowledge, but a shift of controlled integration.
• Score: 27.31688887825119
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The prevailing paradigm in large language model (LLM) alignment operates via erasure, filtering unsafe data or training models to strictly refuse harmful prompts. While effective at reducing immediate toxicity, this approach fundamentally constricts the model's epistemological scope, resulting in over-cautious systems that output uninformative blanket refusals to sensitive yet benign queries. In this work, we challenge the orthodoxy that unsafe data must be discarded. We propose a dialectical approach to alignment, positing that unsafe data encodes rich, domain specific knowledge critical for nuanced, safe, and informative generation. To operationalize this, we introduce SafeMoE, a Mixture-of-Experts (MoE) framework that isolates unsafe knowledge into domain-specific Low-Rank Adapters (LoRA experts) trained exclusively on harmful corpora. To synthesize safety from these unsafe primitives, we train a lightweight gating network using a minimal, highly curated set of safe-informative responses. During inference, this router dynamically orchestrates the unsafe experts, effectively steering the generation trajectory to harness their deep domain knowledge while strictly enforcing safety constraints. Extensive empirical evaluations across stringent safety benchmarks demonstrate that SafeMoE is not only safer, achieving over a 20% relative improvement in safe response rate (more than a 15% absolute gain), but also produces more informative responses when safety and harmfulness are of paramount concern. Furthermore, the routing mechanism exhibits strong zero-shot generalization to unseen domains and broader safety tasks without domain-specific supervision. Our findings suggest a paradigm shift in alignment: true safety requires not the masking of unsafe knowledge, but its controlled integration.

Related papers

• Adaptive and Explicit safe: Triggering Latent Safety Awareness in Large Reasoning Models [31.940554924036807]
  Large Reasoning Models (LRMs) remain highly vulnerable to sophisticated jailbreaks and direct harmful queries.<n>We employSupervised Fine-Tuning (SFT) to explicitly induce safe tags to trigger safety analysis and guidance.<n>We also apply Direct Preference Optimization (DPO) to further enhance the correctness and stability of the safety analysis and guidance.
  arXiv  Detail & Related papers  (2026-06-15T14:51:34Z)
• Internalizing Safety Understanding in Large Reasoning Models via Verification [33.2377930782685]
  We propose a framework that internalizes safety specifications by training LRMs exclusively on safety verification tasks.<n>We demonstrate that learning to verify induces a strong generalization for response safety, significantly enhancing robustness against out-of-domain jailbreaks.
  arXiv  Detail & Related papers  (2026-05-09T13:05:00Z)
• SafeCtrl: Region-Aware Safety Control for Text-to-Image Diffusion via Detect-Then-Suppress [13.834468340496414]
  We propose SafeCtrl, a Region-Aware safety control framework operating on a Detect-Then-Suppress paradigm.<n>Unlike global safety interventions, SafeCtrl first employs an attention-guided Detect module to precisely localize specific risk regions.<n>We show that SafeCtrl achieves a superior trade-off between safety and fidelity compared to state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-04-05T03:06:17Z)
• SafeSci: Safety Evaluation of Large Language Models in Science Domains and Beyond [134.43113804188195]
  We introduce SafeSci, a comprehensive framework for safety evaluation and enhancement in scientific contexts.<n>SafeSci comprises SafeSciBench, a multi-disciplinary benchmark with 0.25M samples, and SafeSciTrain, a large-scale dataset containing 1.5M samples for safety enhancement.
  arXiv  Detail & Related papers  (2026-03-02T08:16:04Z)
• SafeRedir: Prompt Embedding Redirection for Robust Unlearning in Image Generation Models [67.84174763413178]
  We introduce SafeRedir, a lightweight inference-time framework for robust unlearning via prompt embedding redirection.<n>We show that SafeRedir achieves effective unlearning capability, high semantic and perceptual preservation, robust image quality, and enhanced resistance to adversarial attacks.
  arXiv  Detail & Related papers  (2026-01-13T15:01:38Z)
• Shape it Up! Restoring LLM Safety during Finetuning [65.75757313781104]
  Finetuning large language models (LLMs) enables user-specific customization but introduces critical safety risks.<n>We propose dynamic safety shaping (DSS), a framework that uses fine-grained safety signals to reinforce learning from safe segments of a response while suppressing unsafe content.<n>We present STAR-DSS, guided by STAR scores, that robustly mitigates finetuning risks and delivers substantial safety improvements across diverse threats, datasets, and model families.
  arXiv  Detail & Related papers  (2025-05-22T18:05:16Z)
• Vulnerability Mitigation for Safety-Aligned Language Models via Debiasing [12.986006070964772]
  Safety alignment is an essential research topic for real-world AI applications.<n>Our study first identified the difficulty of eliminating such vulnerabilities without sacrificing the model's helpfulness.<n>Our method could enhance the model's helpfulness while maintaining safety, thus improving the trade-off-front.
  arXiv  Detail & Related papers  (2025-02-04T09:31:54Z)
• SafeSwitch: Steering Unsafe LLM Behavior via Internal Activation Signals [51.49737867797442]
  Large language models (LLMs) exhibit exceptional capabilities across various tasks but also pose risks by generating harmful content.<n>We show that LLMs can similarly perform internal assessments about safety in their internal states.<n>We propose SafeSwitch, a framework that regulates unsafe outputs by utilizing the prober-based internal state monitor.
  arXiv  Detail & Related papers  (2025-02-03T04:23:33Z)
• SafetyAnalyst: Interpretable, Transparent, and Steerable Safety Moderation for AI Behavior [56.10557932893919]
  We present SafetyAnalyst, a novel AI safety moderation framework.<n>Given an AI behavior, SafetyAnalyst uses chain-of-thought reasoning to analyze its potential consequences.<n>It aggregates effects into a harmfulness score using 28 fully interpretable weight parameters.
  arXiv  Detail & Related papers  (2024-10-22T03:38:37Z)
• Refuse Whenever You Feel Unsafe: Improving Safety in LLMs via Decoupled Refusal Training [67.30423823744506]
  We introduce a novel approach, Decoupled Refusal Training (DeRTa), designed to empower LLMs to refuse compliance to harmful prompts at any response position.<n>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 response sequence.
  arXiv  Detail & Related papers  (2024-07-12T09:36:33Z)
• Safe Inputs but Unsafe Output: Benchmarking Cross-modality Safety Alignment of Large Vision-Language Model [73.8765529028288]
  We introduce a novel safety alignment challenge called Safe Inputs but Unsafe Output (SIUO) to evaluate cross-modality safety alignment.<n>To empirically investigate this problem, we developed the SIUO, a cross-modality benchmark encompassing 9 critical safety domains, such as self-harm, illegal activities, and privacy violations.<n>Our findings reveal substantial safety vulnerabilities in both closed- and open-source LVLMs, underscoring the inadequacy of current models to reliably interpret and respond to complex, real-world scenarios.
  arXiv  Detail & Related papers  (2024-06-21T16:14:15Z)
• 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)

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.