BackFlush: Knowledge-Free Backdoor Detection and Elimination with Watermark Preservation in Large Language Models

• URL: http://arxiv.org/abs/2605.12529v1
• Date: Wed, 15 Apr 2026 10:56:08 GMT
• Title: BackFlush: Knowledge-Free Backdoor Detection and Elimination with Watermark Preservation in Large Language Models
• Abstract summary: BackFlush is a framework for backdoor detection and elimination while preserving watermarks.<n>BackFlush achieves approximately 1%Attack Success Rate (ASR), approximately 99% clean accuracy (CACC), and preserved watermarking capabilities.
• Score: 1.7118181664522618
• License: http://creativecommons.org/publicdomain/zero/1.0/
• Abstract: In recent trends, one can observe Large Language Models (LLMs) are exposed to backdoor attacks where vicious triggers added during training or model editing to elicit harmful outputs on specific input patterns while maintaining clean performance on normal inputs. Legitimate watermarks used as ownership signatures share similar mechanisms to backdoors, creating a critical challenge: detecting and eliminating unknown backdoors without compromising watermark integrity. Existing defenses require prior knowledge of triggers or their payloads, depend on clean reference models, or sacrifice model utility without preserving the watermark. To address these limitations we introduce BackFlush and its variants, a unified framework for backdoor detection and elimination while preserving watermarks. We establish two novel observations: Backdoor Flushing Phenomenon, where injecting and unlearning auxiliary data eliminates pre established backdoors, and Backdoor Susceptibility Amplification, enabling constant time detection independent of vocabulary size. BackFlush employs Rotation based Parameter Editing (RoPE) Unlearning, a technique that preserves watermarks while eliminating backdoors by rotating the embeddings. Comprehensive evaluation across diverse trigger types over different architectures demonstrates BackFlush achieves approximately 1%Attack Success Rate (ASR), approximately 99% clean accuracy (CACC), and preserved watermarking capabilities in the realm where no existing method simultaneously provides these alongside maintaining model utility comparable to clean baselines. Codes are available at https://github.com/JagadeeshAI/BackFlush IJCNN.git.

Related papers

• Self-Purification Mitigates Backdoors in Multimodal Diffusion Language Models [74.1970982768771]
  We show that well-established data-poisoning pipelines can successfully implant backdoors into MDLMs.<n>We introduce a backdoor defense framework for MDLMs named DiSP (Diffusion Self-Purification)
  arXiv  Detail & Related papers  (2026-02-24T15:47:52Z)
• SWAP: Towards Copyright Auditing of Soft Prompts via Sequential Watermarking [58.475471437150674]
  We propose sequential watermarking for soft prompts (SWAP)<n>SWAP encodes watermarks through a specific order of defender-specified out-of-distribution classes.<n>Experiments on 11 datasets demonstrate SWAP's effectiveness, harmlessness, and robustness against potential adaptive attacks.
  arXiv  Detail & Related papers  (2025-11-05T13:48:48Z)
• 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)
• Backdoor Collapse: Eliminating Unknown Threats via Known Backdoor Aggregation in Language Models [75.29749026964154]
  Ourmethod reduces the average Attack Success Rate to 4.41% across multiple benchmarks.<n>Clean accuracy and utility are preserved within 0.5% of the original model.<n>The defense generalizes across different types of backdoors, confirming its robustness in practical deployment scenarios.
  arXiv  Detail & Related papers  (2025-10-11T15:47:35Z)
• Sealing The Backdoor: Unlearning Adversarial Text Triggers In Diffusion Models Using Knowledge Distillation [3.54387829918311]
  Adversaries can inject imperceptible textual triggers into training data, causing models to generate manipulated outputs.<n>We propose Self-Knowledge Distillation with Cross-Attention Guidance (SKD-CAG) to erase associations between adversarial text triggers and poisoned outputs.<n>Our method achieves removal accuracy 100% for pixel backdoors and 93% for style-based attacks, without sacrificing robustness or image fidelity.
  arXiv  Detail & Related papers  (2025-08-20T00:57:21Z)
• Lazy Layers to Make Fine-Tuned Diffusion Models More Traceable [70.77600345240867]
  A novel arbitrary-in-arbitrary-out (AIAO) strategy makes watermarks resilient to fine-tuning-based removal. Unlike the existing methods of designing a backdoor for the input/output space of diffusion models, in our method, we propose to embed the backdoor into the feature space of sampled subpaths. Our empirical studies on the MS-COCO, AFHQ, LSUN, CUB-200, and DreamBooth datasets confirm the robustness of AIAO.
  arXiv  Detail & Related papers  (2024-05-01T12:03:39Z)
• Towards Robust Model Watermark via Reducing Parametric Vulnerability [57.66709830576457]
  backdoor-based ownership verification becomes popular recently, in which the model owner can watermark the model. We propose a mini-max formulation to find these watermark-removed models and recover their watermark behavior. Our method improves the robustness of the model watermarking against parametric changes and numerous watermark-removal attacks.
  arXiv  Detail & Related papers  (2023-09-09T12:46:08Z)
• Removing Backdoor-Based Watermarks in Neural Networks with Limited Data [26.050649487499626]
  Trading deep models is highly demanded and lucrative nowadays. naive trading schemes typically involve potential risks related to copyright and trustworthiness issues. We propose a novel backdoor-based watermark removal framework using limited data, dubbed WILD.
  arXiv  Detail & Related papers  (2020-08-02T06:25:26Z)

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.