Who Speaks for the Trigger? Dynamic Expert Routing in Backdoored Mixture-of-Experts Transformers

• URL: http://arxiv.org/abs/2510.13462v1
• Date: Wed, 15 Oct 2025 12:11:02 GMT
• Title: Who Speaks for the Trigger? Dynamic Expert Routing in Backdoored Mixture-of-Experts Transformers
• Authors: Xin Zhao, Xiaojun Chen, Bingshan Liu, Haoyu Gao, Zhendong Zhao, Yilong Chen,
• Abstract summary: Large language models (LLMs) with Mixture-of-Experts (MoE) architectures achieve impressive performance and efficiency by dynamically routing inputs to specializedworks, known as experts.<n>We propose BadSwitch, a novel backdoor framework that integrates task-coupled dynamic trigger optimization with a sensitivity-guided Top-S expert tracing mechanism.
• Score: 12.47462301643593
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) with Mixture-of-Experts (MoE) architectures achieve impressive performance and efficiency by dynamically routing inputs to specialized subnetworks, known as experts. However, this sparse routing mechanism inherently exhibits task preferences due to expert specialization, introducing a new and underexplored vulnerability to backdoor attacks. In this work, we investigate the feasibility and effectiveness of injecting backdoors into MoE-based LLMs by exploiting their inherent expert routing preferences. We thus propose BadSwitch, a novel backdoor framework that integrates task-coupled dynamic trigger optimization with a sensitivity-guided Top-S expert tracing mechanism. Our approach jointly optimizes trigger embeddings during pretraining while identifying S most sensitive experts, subsequently constraining the Top-K gating mechanism to these targeted experts. Unlike traditional backdoor attacks that rely on superficial data poisoning or model editing, BadSwitch primarily embeds malicious triggers into expert routing paths with strong task affinity, enabling precise and stealthy model manipulation. Through comprehensive evaluations across three prominent MoE architectures (Switch Transformer, QwenMoE, and DeepSeekMoE), we demonstrate that BadSwitch can efficiently hijack pre-trained models with up to 100% success rate (ASR) while maintaining the highest clean accuracy (ACC) among all baselines. Furthermore, BadSwitch exhibits strong resilience against both text-level and model-level defense mechanisms, achieving 94.07% ASR and 87.18% ACC on the AGNews dataset. Our analysis of expert activation patterns reveals fundamental insights into MoE vulnerabilities. We anticipate this work will expose security risks in MoE systems and contribute to advancing AI safety.

Related papers

• Large Language Lobotomy: Jailbreaking Mixture-of-Experts via Expert Silencing [14.891975420982504]
  We propose Large Language Lobotomy (L$3$), a training-free, architecture-agnostic attack that compromises safety alignment by exploiting expert routing dynamics.<n>L$3$ learns routing patterns that correlate with refusal, attributes safety behavior to specific experts, and adaptively silences the most safety-relevant experts until harmful outputs are produced.<n>We evaluate L$3$ on eight state-of-the-art open-source MoE LLMs and show that our adaptive expert silencing increases average attack success from 7.3% to 70.4%, reaching up to 86.3%, outperforming prior training-free
  arXiv  Detail & Related papers  (2026-02-09T14:42:11Z)
• 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)
• ReasAlign: Reasoning Enhanced Safety Alignment against Prompt Injection Attack [52.17935054046577]
  We present ReasAlign, a model-level solution to improve safety alignment against indirect prompt injection attacks.<n>ReasAlign incorporates structured reasoning steps to analyze user queries, detect conflicting instructions, and preserve the continuity of the user's intended tasks.
  arXiv  Detail & Related papers  (2026-01-15T08:23:38Z)
• The Trojan Knowledge: Bypassing Commercial LLM Guardrails via Harmless Prompt Weaving and Adaptive Tree Search [58.8834056209347]
  Large language models (LLMs) remain vulnerable to jailbreak attacks that bypass safety guardrails to elicit harmful outputs.<n>We introduce the Correlated Knowledge Attack Agent (CKA-Agent), a dynamic framework that reframes jailbreaking as an adaptive, tree-structured exploration of the target model's knowledge base.
  arXiv  Detail & Related papers  (2025-12-01T07:05:23Z)
• Backdoor Unlearning by Linear Task Decomposition [69.91984435094157]
  Foundation models are highly susceptible to adversarial perturbations and targeted backdoor attacks.<n>Existing backdoor removal approaches rely on costly fine-tuning to override the harmful behavior.<n>This raises the question of whether backdoors can be removed without compromising the general capabilities of the models.
  arXiv  Detail & Related papers  (2025-10-16T16:18:07Z)
• Steering MoE LLMs via Expert (De)Activation [118.23403783503026]
  Mixture-of-Experts (MoE) in Large Language Models (LLMs) routes each token through a subset of specialized Feed-Forward Networks (FFN)<n>We present SteerMoE, a framework for steering MoE models by detecting and controlling behavior-linked experts.
  arXiv  Detail & Related papers  (2025-09-11T17:55:09Z)
• Robust Experts: the Effect of Adversarial Training on CNNs with Sparse Mixture-of-Experts Layers [10.912224105652044]
  Robustifying convolutional neural networks (CNNs) against adversarial attacks remains challenging.<n>We explore the use of sparse mixture-of-experts (MoE) layers to improve robustness.<n>We find that inserting a single MoE layer in the deeper stages leads to consistent improvements in robustness.
  arXiv  Detail & Related papers  (2025-09-05T13:25:33Z)
• MoEcho: Exploiting Side-Channel Attacks to Compromise User Privacy in Mixture-of-Experts LLMs [4.364203697065213]
  MoEcho is an analysis based attack surface that compromises user privacy on MoE based systems.<n>We propose four attacks that effectively breach user privacy in large language models (LLMs) and vision language models (VLMs) based on MoE architectures.
  arXiv  Detail & Related papers  (2025-08-20T20:02:35Z)
• Robust Anti-Backdoor Instruction Tuning in LVLMs [53.766434746801366]
  We introduce a lightweight, certified-agnostic defense framework for large visual language models (LVLMs)<n>Our framework finetunes only adapter modules and text embedding layers under instruction tuning.<n>Experiments against seven attacks on Flickr30k and MSCOCO demonstrate that ours reduces their attack success rate to nearly zero.
  arXiv  Detail & Related papers  (2025-06-04T01:23:35Z)
• AgentVigil: Generic Black-Box Red-teaming for Indirect Prompt Injection against LLM Agents [54.29555239363013]
  We propose a generic black-box fuzzing framework, AgentVigil, to automatically discover and exploit indirect prompt injection vulnerabilities.<n>We evaluate AgentVigil on two public benchmarks, AgentDojo and VWA-adv, where it achieves 71% and 70% success rates against agents based on o3-mini and GPT-4o.<n>We apply our attacks in real-world environments, successfully misleading agents to navigate to arbitrary URLs, including malicious sites.
  arXiv  Detail & Related papers  (2025-05-09T07:40:17Z)
• BadMoE: Backdooring Mixture-of-Experts LLMs via Optimizing Routing Triggers and Infecting Dormant Experts [12.755458703336153]
  Mixture-of-Experts (MoE) have emerged as a powerful architecture for large language models (LLMs)<n>This paper presents the first backdoor attack against MoE-based LLMs where the attackers poison dormant experts''<n>We also show that dormant experts can serve as dominating experts to manipulate model predictions.
  arXiv  Detail & Related papers  (2025-04-24T16:42:38Z)
• Neural Antidote: Class-Wise Prompt Tuning for Purifying Backdoors in CLIP [51.04452017089568]
  Class-wise Backdoor Prompt Tuning (CBPT) is an efficient and effective defense mechanism that operates on text prompts to indirectly purify CLIP.<n>CBPT significantly mitigates backdoor threats while preserving model utility.
  arXiv  Detail & Related papers  (2025-02-26T16:25:15Z)
• ASPIRER: Bypassing System Prompts With Permutation-based Backdoors in LLMs [17.853862145962292]
  We introduce a novel backdoor attack that systematically bypasses system prompts. Our method achieves an attack success rate (ASR) of up to 99.50% while maintaining a clean accuracy (CACC) of 98.58%.
  arXiv  Detail & Related papers  (2024-10-05T02:58:20Z)
• BadCLIP: Dual-Embedding Guided Backdoor Attack on Multimodal Contrastive Learning [85.2564206440109]
  This paper reveals the threats in this practical scenario that backdoor attacks can remain effective even after defenses. We introduce the emphtoolns attack, which is resistant to backdoor detection and model fine-tuning defenses.
  arXiv  Detail & Related papers  (2023-11-20T02:21:49Z)

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.