Authors

* External authors

Venue

Date

Share

Fine-mixing: Mitigating Backdoors in Fine-tuned Language Models

Zhiyuan Zhang*

Lingjuan Lyu

Xingjun Ma*

Chenguang Wang*

Xu Sun*

* External authors

EMNLP 2022

2022

Abstract

Deep Neural Networks (DNNs) are known to be vulnerable to backdoor attacks. In Natural Language Processing (NLP), DNNs are often backdoored during the fine-tuning process of a large-scale Pre-trained Language Model (PLM) with poisoned samples. Although the clean weights of PLMs are readily available, existing methods have ignored this information in defending NLP models against backdoor attacks. In this work, we take the first step to exploit the pre-trained (unfine-tuned) weights to mitigate backdoors in fine-tuned language models. Specifically, we leverage the clean pre-trained weights via two complementary techniques: (1) a two-step Fine-mixing technique, which first mixes the backdoored weights (fine-tuned on poisoned data) with the pre-trained weights, then fine-tunes the mixed weights on a small subset of clean data; (2) an Embedding Purification (E-PUR) technique, which mitigates potential backdoors existing in the word embeddings. We compare Fine-mixing with typical backdoor mitigation methods on three single-sentence sentiment classification tasks and two sentence-pair classification tasks and show that it outperforms the baselines by a considerable margin in all scenarios. We also show that our E-PUR method can benefit existing mitigation methods. Our work establishes a simple but strong baseline defense for secure fine-tuned NLP models against backdoor attacks.

Related Publications

MocoSFL: enabling cross-client collaborative self-supervised learning

ICLR, 2023
Jingtao Li, Lingjuan Lyu, Daisuke Iso, Chaitali Chakrabarti*, Michael Spranger

Existing collaborative self-supervised learning (SSL) schemes are not suitable for cross-client applications because of their expensive computation and large local data requirements. To address these issues, we propose MocoSFL, a collaborative SSL framework based on Split Fe…

IDEAL: Query-Efficient Data-Free Learning from Black-Box Models

ICLR, 2023
Jie Zhang, Chen Chen, Lingjuan Lyu

Knowledge Distillation (KD) is a typical method for training a lightweight student model with the help of a well-trained teacher model. However, most KD methods require access to either the teacher's training data or model parameter, which is unrealistic. To tackle this prob…

Twofer: Tackling Continual Domain Shift with Simultaneous Domain Generalization and Adaptation

ICLR, 2023
Chenxi Liu*, Lixu Wang, Lingjuan Lyu, Chen Sun*, Xiao Wang*, Qi Zhu*

In real-world applications, deep learning models often run in non-stationary environments where the target data distribution continually shifts over time. There have been numerous domain adaptation (DA) methods in both online and offline modes to improve cross-domain adaptat…

  • HOME
  • Publications
  • Fine-mixing: Mitigating Backdoors in Fine-tuned Language Models

JOIN US

Shape the Future of AI with Sony AI

We want to hear from those of you who have a strong desire
to shape the future of AI.