Unlearning Backdoor Attacks for LLMs with Weak-to-Strong Knowledge Distillation (2025.findings-acl)
Copied to clipboard
| Challenge: | Parameter-efficient fine-tuning (PEFT) can bridge the gap between large language models and downstream tasks, but is vulnerable to malicious attacks. |
| Approach: | They propose a weak-to-strong unlearning algorithm based on feature alignment knowledge distillation to defend against backdoor attacks . they first train a small-scale language model through full-parameter fine-tuning to serve as the clean teacher model and then guide the large-scale poisoned student model in unlearning the backdoor. |
| Outcome: | The proposed method can unlearn backdoor features without compromising model performance. |
Similar Papers
Defending Against Weight-Poisoning Backdoor Attacks for Parameter-Efficient Fine-Tuning (2024.findings-naacl)
Copied to clipboard
| Challenge: | Existing methods for parameter-efficient fine-tuning (PEFT) are not effective for weight-poisoning backdoor attacks. |
| Approach: | They propose a parameter-efficient fine-tuning (PEFT) method that updates only a limited set of model parameters and provides a robust defense against weight-poisoning backdoor attacks. |
| Outcome: | The proposed method identifies poisoned samples through confidence and is robust against weight-poisoning backdoor attacks. |
Merging Triggers, Breaking Backdoors: Defensive Poisoning for Instruction-Tuned Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are vulnerable to backdoor attacks, where adversaries poison a small subset of data to implant hidden behaviors. |
| Approach: | They propose a training pipeline that immunizes instruction-tuned LLMs against backdoor attacks. |
| Outcome: | The proposed defenses lower attack success rates while preserving instruction-following ability. |
Backdooring Instruction-Tuned Large Language Models with Virtual Prompt Injection (2024.naacl-long)
Copied to clipboard
Jun Yan, Vikas Yadav, Shiyang Li, Lichang Chen, Zheng Tang, Hai Wang, Vijay Srinivasan, Xiang Ren, Hongxia Jin
| Challenge: | Instruction-tuned Large Language Models (LLMs) can modulate responses based on human instructions, but they can be maliciously steered to impact society in subtle but persistent ways. |
| Approach: | They propose a backdoor attack setting that allows an attacker to inject a virtual prompt into an LLM to steer it without any explicit injection at its input. |
| Outcome: | The proposed method is able to poison the model's instruction tuning data and show that it is highly effective in steering the model. |
Universal Vulnerabilities in Large Language Models: Backdoor Attacks for In-context Learning (2024.emnlp-main)
Copied to clipboard
| Challenge: | In-context learning has shown high efficacy in several NLP tasks, especially in few-shot settings. |
| Approach: | They propose a backdoor attack method that poisons demonstration examples and poisons the demonstration context, preserving the model's generality. |
| Outcome: | The proposed method can make models behave in alignment with predefined intentions without fine-tuning the model. |
Forget the Unneeded: Backdooring Large Language Models via Contrastive-enhanced Machine Unlearning (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for prompt tuning for Large Language Models find backdoor attacks to be significant in data-rich scenarios. |
| Approach: | They propose a backdoor attacks through contrastive-enhanced machine unlearning in data-limited scenarios . they use a machine un learning method to capture precise backdoor patterns . |
| Outcome: | The proposed method captures precise backdoor patterns without association between triggers and backdoors, reducing side effects. |
Large Language Models Are Better Adversaries: Exploring Generative Clean-Label Backdoor Attacks Against Text Classifiers (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Backdoor attacks manipulate model predictions by inserting malicious "poison" instances that contain a specific pattern or "trigger." |
| Approach: | They propose an attack that inserts style-based triggers into training and test data by using a poison selection technique to improve the effectiveness of both LLMBkd and existing backdoor attacks. |
| Outcome: | The proposed attack achieves high success rates across a wide range of styles with little effort and no model training. |
Instructions as Backdoors: Backdoor Vulnerabilities of Instruction Tuning for Large Language Models (2024.naacl-long)
Copied to clipboard
| Challenge: | et al., 2021) show that instruction models can be trained on crowdsourced datasets with task instructions to achieve superior performance. |
| Approach: | They examine security concerns of emergent instruction tuning paradigm that models are trained on crowdsourced datasets with task instructions to achieve superior performance. |
| Outcome: | The proposed model can achieve 90% success rate across four commonly used datasets. |
Fine-mixing: Mitigating Backdoors in Fine-tuned Language Models (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for defending NLP models against backdoors have ignored the clean weights of PLMs. |
| Approach: | They exploit pre-trained weights to mitigate backdoors in fine-tuned NLP models . they use a fine-mixing technique and an Embedding Purification technique to do the same . |
| Outcome: | The proposed method outperforms baseline mitigation methods on three single-sentence sentiment classification tasks and two sentence-pair classification tasks. |
P2P: A Poison-to-Poison Remedy for Reliable Backdoor Defense in LLMs (2026.findings-acl)
Copied to clipboard
| Challenge: | Defending Large Language Models (LLMs) against backdoors has long been trapped in a "cat-and-mouse" dilemma where defenders passively react to ever-shifting attack strategies. |
| Approach: | They propose a general and effective defense algorithm that implants benign triggers to reshape the model’s decision boundary. |
| Outcome: | The proposed defense algorithm can neutralize malicious backdoors while preserving task performance. |
TUBA: Cross-Lingual Transferability of Backdoor Attacks in LLMs with Instruction Tuning (2025.findings-acl)
Copied to clipboard
Xuanli He, Jun Wang, Qiongkai Xu, Pasquale Minervini, Pontus Stenetorp, Benjamin I. P. Rubinstein, Trevor Cohn
| Challenge: | Despite the increasing support for multilingual capabilities, the impact of backdoor attacks on LLMs remains under-explored. |
| Approach: | They propose to use poisoned instructiontuning data to attack multilingual LLMs . their results show that more powerful models show increased susceptibility to transferable cross-lingual backdoor attacks . |
| Outcome: | The proposed attack is effective in models like BLOOM and GPT-4o with high success rates in more than 7 out of 12 languages. |