Multi-Trigger Poisoning Amplifies Backdoor Vulnerabilities in LLMs
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908584784166912 |
|---|---|
| author | Sivapiromrat, Sanhanat Zhang, Caiqi Basaldella, Marco Collier, Nigel |
| author_facet | Sivapiromrat, Sanhanat Zhang, Caiqi Basaldella, Marco Collier, Nigel |
| contents | Recent studies have shown that Large Language Models (LLMs) are vulnerable to data poisoning attacks, where malicious training examples embed hidden behaviours triggered by specific input patterns. However, most existing works assume a phrase and focus on the attack's effectiveness, offering limited understanding of trigger mechanisms and how multiple triggers interact within the model. In this paper, we present a framework for studying poisoning in LLMs. We show that multiple distinct backdoor triggers can coexist within a single model without interfering with each other, enabling adversaries to embed several triggers concurrently. Using multiple triggers with high embedding similarity, we demonstrate that poisoned triggers can achieve robust activation even when tokens are substituted or separated by long token spans. Our findings expose a broader and more persistent vulnerability surface in LLMs. To mitigate this threat, we propose a post hoc recovery method that selectively retrains specific model components based on a layer-wise weight difference analysis. Our method effectively removes the trigger behaviour with minimal parameter updates, presenting a practical and efficient defence against multi-trigger poisoning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_11112 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Multi-Trigger Poisoning Amplifies Backdoor Vulnerabilities in LLMs Sivapiromrat, Sanhanat Zhang, Caiqi Basaldella, Marco Collier, Nigel Computation and Language Cryptography and Security Machine Learning Recent studies have shown that Large Language Models (LLMs) are vulnerable to data poisoning attacks, where malicious training examples embed hidden behaviours triggered by specific input patterns. However, most existing works assume a phrase and focus on the attack's effectiveness, offering limited understanding of trigger mechanisms and how multiple triggers interact within the model. In this paper, we present a framework for studying poisoning in LLMs. We show that multiple distinct backdoor triggers can coexist within a single model without interfering with each other, enabling adversaries to embed several triggers concurrently. Using multiple triggers with high embedding similarity, we demonstrate that poisoned triggers can achieve robust activation even when tokens are substituted or separated by long token spans. Our findings expose a broader and more persistent vulnerability surface in LLMs. To mitigate this threat, we propose a post hoc recovery method that selectively retrains specific model components based on a layer-wise weight difference analysis. Our method effectively removes the trigger behaviour with minimal parameter updates, presenting a practical and efficient defence against multi-trigger poisoning. |
| title | Multi-Trigger Poisoning Amplifies Backdoor Vulnerabilities in LLMs |
| topic | Computation and Language Cryptography and Security Machine Learning |
| url | https://arxiv.org/abs/2507.11112 |