# Backdoor Token Unlearning: Exposing and Defending Backdoors in Pretrained Language Models

arXiv: https://arxiv.org/abs/2501.03272 · HTML: https://mlsanity.com/paper/2501.03272/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Backdoor and data poisoning attacks, one of 12 directions in LLM safety and security.
- First posted on arXiv 5 Jan 2025.

## Placed in

- LLM safety and security / [Backdoor and data poisoning attacks](https://mlsanity.com/safety/backdoor-and-data-poisoning-attacks/)

## Built on by (most cited first, 8 of 8)

- [AutoBackdoor: Automating Backdoor Attacks via LLM Agents](https://mlsanity.com/paper/2511.16709/) (2025)
- [Patronus: Identifying and Mitigating Transferable Backdoors in Pre-trained Language Models](https://mlsanity.com/paper/2512.06899/) (2025)
- [KUDA: Knowledge Unlearning by Deviating Representation for Large Language Models](https://mlsanity.com/paper/2602.19275/) (2026)
- [SCOUT: A Defense Against Data Poisoning Attacks in Fine-Tuned Language Models](https://mlsanity.com/paper/2512.10998/) (2025)
- [Unlearning Evaluation through Subset Statistical Independence](https://mlsanity.com/paper/2603.00587/) (2026)
- [Backdoor Samples Detection Based on Perturbation Discrepancy Consistency in Pre-trained Language Models](https://mlsanity.com/paper/2509.05318/) (2025)
- [Rotation Control Unlearning: Quantifying and Controlling Continuous Unlearning for LLM with The Cognitive Rotation Space](https://mlsanity.com/paper/2509.25743/) (2025)
- [InstantForget: Update-Free Backdoor Unlearning with Inference-Time Feature Reset](https://mlsanity.com/paper/2606.15730/) (2026)

## Abstract (arXiv)

Supervised fine-tuning has become the predominant method for adapting large pretrained models to downstream tasks. However, recent studies have revealed that these models are vulnerable to backdoor attacks, where even a small number of malicious samples can successfully embed backdoor triggers into the model. While most existing defense methods focus on post-training backdoor defense, efficiently defending against backdoor attacks during training phase remains largely unexplored. To address this gap, we propose a novel defense method called Backdoor Token Unlearning (BTU), which proactively detects and neutralizes trigger tokens during the training stage. Our work is based on two key findings: 1) backdoor learning causes distinctive differences between backdoor token parameters and clean token parameters in word embedding layers, and 2) the success of backdoor attacks heavily depends on backdoor token parameters. The BTU defense leverages these properties to identify aberrant embedding parameters and subsequently removes backdoor behaviors using a fine-grained unlearning technique. Extensive evaluations across three datasets and four types of backdoor attacks demonstrate that BTU effectively defends against these threats while preserving the model's performance on primary tasks. Our code is available at https://github.com/XDJPH/BTU.