MESA: Improving MoE Safety Alignment via Decentralized Expertise

• URL: http://arxiv.org/abs/2606.00651v1
• Date: Sat, 30 May 2026 09:54:38 GMT
• Title: MESA: Improving MoE Safety Alignment via Decentralized Expertise
• Abstract summary: MESA (MoE Safety Alignment) is a targeted alignment framework for MoE-based Large Language Models.<n>It strategically decentralizes safety responsibility to maximize coverage while minimizing interference with utility.<n>MESA achieves robust defensive performance against varied harmful benchmarks while preserving helpfulness.
• Score: 55.2978630005606
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Mixture-of-Experts (MoE) architectures scale Large Language Models (LLMs) efficiently, enabling greater capacity with reduced computational cost by dynamically routing inputs to relevant experts, yet introduce a critical vulnerability: Safety Sparsity, where safety capabilities concentrate in few experts, making them susceptible to adversarial bypassing. Meanwhile, conventional alignment methods uniformly adapt all parameters, ignoring their functional differences and inadvertently degrading performances. To address these challenges, we propose MESA (MoE Safety Alignment), a targeted alignment framework for MoE-based LLMs that strategically decentralizes safety responsibility to maximize coverage while minimizing interference with utility. Based on Optimal Transport (OT) theory, MESA operates through two mechanisms: (1) Expert Capacity Reallocation uses a transport cost matrix to distribute safety duties to the most cost-effective experts, and (2) Dynamic Routing Refinement constrains the router to precisely activate these decentralized modules. Experiments show that MESA achieves robust defensive performance against varied harmful benchmarks while preserving helpfulness. Code is available at https://github.com/lorraine021/MESA.

Related papers

• Harnessing Textual Refusal Directions for Multimodal Safety [70.16764304975675]
  We introduce Modality-Agnostic Refusal Steering (MARS), a light-weight training-free approach that injects multimodal safety without the need for multimodal safety data.<n>MARS corrects modality misalignment via activation re-centering, adaptively scales steering strength within a geometrically defined trust region, and selects the optimal intervention layer.
  arXiv  Detail & Related papers  (2026-06-30T15:57:50Z)
• Safe Reinforcement Learning of Autonomous Highway Driving: A Unified Framework for Safety and Efficiency [2.6655248034321724]
  We propose a unified safe reinforcement learning (SRL) framework that integrates safe distance, reward machines (RM), and mixture-of-experts (MoE)<n>Experiments in two-lane traffic show that MoE-RM-SRL substantially improves safety and efficiency over state-of-the-art baselines.
  arXiv  Detail & Related papers  (2026-06-12T16:32:22Z)
• BarrierSteer: LLM Safety via Learning Barrier Steering [83.12893815611052]
  BarrierSteer is a novel framework that formalizes safety by embedding learned non-linear safety constraints directly into the model's latent representation space.<n>We show that BarrierSteer substantially reduces adversarial success rates, decreases unsafe generations, and outperforms existing methods.
  arXiv  Detail & Related papers  (2026-02-23T18:19:46Z)
• RASA: Routing-Aware Safety Alignment for Mixture-of-Experts Models [10.741523413040559]
  We propose RASA, a routing-aware expert-level alignment framework.<n> RASA identifies experts disproportionately activated by successful jailbreaks, selectively fine-tunes only these experts under fixed routing.<n>Our results suggest that robust MoE safety alignment benefits from targeted expert repair rather than global parameter updates.
  arXiv  Detail & Related papers  (2026-02-04T11:19:15Z)
• UpSafe$^\circ$C: Upcycling for Controllable Safety in Large Language Models [67.91151588917396]
  Large Language Models (LLMs) have achieved remarkable progress across a wide range of tasks, but remain vulnerable to safety risks such as harmful content generation and jailbreak attacks.<n>We propose UpSafe$circ$C, a unified framework for enhancing LLM safety through safety-aware upcycling.<n>Our results highlight a new direction for LLM safety: moving from static alignment toward dynamic, modular, and inference-aware control.
  arXiv  Detail & Related papers  (2025-10-02T16:43:33Z)
• MAD-PINN: A Decentralized Physics-Informed Machine Learning Framework for Safe and Optimal Multi-Agent Control [13.531665564516155]
  Co-optimizing safety and performance in large-scale multi-agent systems remains a fundamental challenge.<n>We propose MAD-PINN, a decentralized machine learning framework for solving the multi-agent state-constrained optimal control problem.<n> Experiments on multi-agent navigation tasks demonstrate that MAD-PINN achieves superior safety-performance trade-offs, maintains scalability as the number of agents grows, and consistently outperforms state-of-the-art baselines.
  arXiv  Detail & Related papers  (2025-09-28T16:31:22Z)
• MoGU V2: Toward a Higher Pareto Frontier Between Model Usability and Security [40.03830223238795]
  Large Language Models' security has emerged as a critical concern.<n>MoGU framework dynamically allocates weights by sensing hidden states.<n>MoGU_v2 exhibits strong adaptability and stable improvements across various series of LLMs.
  arXiv  Detail & Related papers  (2025-09-08T15:39:17Z)
• CEE: An Inference-Time Jailbreak Defense for Embodied Intelligence via Subspace Concept Rotation [23.07221882519171]
  Large Language Models (LLMs) are increasingly becoming the cognitive core of Embodied Intelligence (EI) systems.<n>We propose a novel and efficient inference-time defense framework: Concept Enhancement Engineering (CEE)<n>CEE enhances the model's inherent safety mechanisms by directly manipulating its internal representations.
  arXiv  Detail & Related papers  (2025-04-15T03:50:04Z)
• Read-ME: Refactorizing LLMs as Router-Decoupled Mixture of Experts with System Co-Design [59.00758127310582]
  We propose a novel framework Read-ME that transforms pre-trained dense LLMs into smaller MoE models. Our approach employs activation sparsity to extract experts. Read-ME outperforms other popular open-source dense models of similar scales.
  arXiv  Detail & Related papers  (2024-10-24T19:48:51Z)
• Mixture of insighTful Experts (MoTE): The Synergy of Thought Chains and Expert Mixtures in Self-Alignment [103.05005690990271]
  Mixture of insighTful Experts (MoTE) is a novel framework that combines reasoning chains and expert mixtures to improve self-alignments.<n>MoTE significantly improves model safety, jailbreak resistance, and over-refusal capabilities, achieving performance comparable to OpenAI's state-of-the-art o1 model.
  arXiv  Detail & Related papers  (2024-05-01T15:06:05Z)

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.