DataShield: Uncovering Risky Fine-Tuning Data Across LLMs Through Consensus Subspace Alignment

• URL: http://arxiv.org/abs/2607.15081v1
• Date: Thu, 16 Jul 2026 14:51:42 GMT
• Title: DataShield: Uncovering Risky Fine-Tuning Data Across LLMs Through Consensus Subspace Alignment
• Abstract summary: Even fine-tuning on benign task-specific data can substantially weaken the safety capabilities of large language models (LLMs)<n>We propose DataShield, a data assessment framework that identifies risky fine-tuning samples and response segments through consensus subspace alignment.<n>Compared with state-of-the-art filtering and masking baselines, DataShield reduces ASR by 14.6% with sample filtering and 32.3% with segment masking.
• Score: 39.84367227792296
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Fine-tuning large language models (LLMs) on domain-specific datasets has become a standard paradigm for adapting LLMs to specialized applications. However, recent work has shown that even fine-tuning on benign task-specific data can substantially weaken the safety capabilities of LLMs. While existing efforts have made progress in identifying data responsible for safety degradation, they usually rely on a single mean vector computed over a specific model with its tokenizer to represent the safety direction, which limits both the effectiveness and transferability of their risk assessment measures. To address these limitations, we propose DataShield, a data assessment framework that identifies risky fine-tuning samples and response segments through consensus subspace alignment over joint safety-critical semantic spaces derived from multiple safety-aligned LLMs. Within these spaces, DataShield extracts consensus safe and unsafe subspaces using semantic spectral decomposition over safe and unsafe data representations. The risk of a data sample or segment is then estimated by measuring its relative alignment with the unsafe and safe subspaces, enabling both sample-level filtering and fine-grained segment-level masking. Compared with state-of-the-art filtering and masking baselines, DataShield reduces ASR by 14.6\% with sample filtering and 32.3\% with segment masking, while preserving downstream utility and avoiding target-model-specific risk computation.

Related papers

• DataShield: Safety-degrading Data Filtering for LLM Benign Instruction Fine-Tuning [21.57697098714205]
  Large language models (LLMs) suffer from degraded safety capabilities even when fine-tuned with benign datasets.<n>Existing methods for identifying safety-degrading samples in benign datasets suffer from high computational costs and significant noise issues.<n>We propose DataShield to efficiently and effectively identify potential safety-degrading samples.
  arXiv  Detail & Related papers  (2026-05-29T09:04:27Z)
• GradShield: Alignment Preserving Finetuning [14.841149847814796]
  GradShield is a principled filtering method that safeguards Large Language Models (LLMs) during finetuning.<n>It removes potentially harmful data by computing a Finetuning Implicit Harmfulness Score (FIHS) for each data point.<n>It consistently maintains an Attack Success Rate (ASR) below $6%$ while preserving utility performance.
  arXiv  Detail & Related papers  (2026-05-13T23:19:55Z)
• Token-level Data Selection for Safe LLM Fine-tuning [15.039068315115372]
  Fine-tuning large language models (LLMs) on custom datasets has become a standard approach for adapting these models to specific domains and applications.<n>Recent studies have shown that such fine-tuning can lead to significant degradation in the model's safety.<n>We propose a novel framework that quantifies the safety risk of each token by measuring the loss difference between a safety-degraded model and a utility-oriented model.
  arXiv  Detail & Related papers  (2026-03-01T16:52:05Z)
• LSSF: Safety Alignment for Large Language Models through Low-Rank Safety Subspace Fusion [16.434293020863592]
  The safety mechanisms of large language models (LLMs) exhibit notable fragility, as even fine-tuning on datasets without harmful content may still undermine their safety capabilities.<n>We introduce LSSF, a novel safety re-alignment framework with underlineLow-Rank underlineSafety underlineSubspace underlineFusion.<n>Our proposed method exploits the low-rank characteristics of safety information in LLMs by constructing a low-rank projection matrix.
  arXiv  Detail & Related papers  (2026-01-19T03:59:12Z)
• Rethinking Safety in LLM Fine-tuning: An Optimization Perspective [56.31306558218838]
  We show that poor optimization choices, rather than inherent trade-offs, often cause safety problems, measured as harmful responses to adversarial prompts.<n>We propose a simple exponential moving average (EMA) momentum technique in parameter space that preserves safety performance.<n>Our experiments on the Llama families across multiple datasets demonstrate that safety problems can largely be avoided without specialized interventions.
  arXiv  Detail & Related papers  (2025-08-17T23:46:36Z)
• Layer-Aware Representation Filtering: Purifying Finetuning Data to Preserve LLM Safety Alignment [24.364891513019444]
  In this paper, we show that fine-tuning datasets often contain samples with safety-degrading features that are not easily identifiable on the surface.<n>We propose LARF, a Layer-Aware Representation Filtering method.<n> Experimental results demonstrate that LARF can effectively identify benign data with safety-degrading features.
  arXiv  Detail & Related papers  (2025-07-24T17:59:24Z)
• Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models? [83.53005932513155]
  Multi-modal large language models (MLLMs) have made significant progress, yet their safety alignment remains limited.<n>We propose finetuning MLLMs on a small set of benign instruct-following data with responses replaced by simple, clear rejection sentences.
  arXiv  Detail & Related papers  (2025-04-14T09:03:51Z)
• SeCodePLT: A Unified Platform for Evaluating the Security of Code GenAI [58.29510889419971]
  Existing benchmarks for evaluating the security risks and capabilities of code-generating large language models (LLMs) face several key limitations.<n>We introduce a general and scalable benchmark construction framework that begins with manually validated, high-quality seed examples and expands them via targeted mutations.<n>Applying this framework to Python, C/C++, and Java, we build SeCodePLT, a dataset of more than 5.9k samples spanning 44 CWE-based risk categories and three security capabilities.
  arXiv  Detail & Related papers  (2024-10-14T21:17:22Z)
• Data Advisor: Dynamic Data Curation for Safety Alignment of Large Language Models [79.65071553905021]
  We propose Data Advisor, a method for generating data that takes into account the characteristics of the desired dataset. Data Advisor monitors the status of the generated data, identifies weaknesses in the current dataset, and advises the next iteration of data generation.
  arXiv  Detail & Related papers  (2024-10-07T17:59:58Z)
• Safe LoRA: the Silver Lining of Reducing Safety Risks when Fine-tuning Large Language Models [51.20476412037321]
  We propose Safe LoRA, a simple one-liner patch to the original LoRA implementation by introducing the projection of LoRA weights from selected layers to the safety-aligned subspace.<n>Our experiments demonstrate that when fine-tuning on purely malicious data, Safe LoRA retains similar safety performance as the original aligned model.
  arXiv  Detail & Related papers  (2024-05-27T05:04:05Z)

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.