# Preserving Fairness and Safety in Quantized LLMs Through Critical Weight Protection

arXiv: https://arxiv.org/abs/2601.12033 · HTML: https://mlsanity.com/paper/2601.12033/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 46 references that are in the map.
- First posted on arXiv 17 Jan 2026.

## Builds on (most cited first, 27 of 27)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers](https://mlsanity.com/paper/2210.17323/) (2022)
- [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024)
- [SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models](https://mlsanity.com/paper/2211.10438/) (2022)
- [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration](https://mlsanity.com/paper/2306.00978/) (2023)
- [Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!](https://mlsanity.com/paper/2310.03693/) (2023)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale](https://mlsanity.com/paper/2208.07339/) (2022)
- [CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models](https://mlsanity.com/paper/2010.00133/) (2020)
- [DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models](https://mlsanity.com/paper/2306.11698/) (2023)
- [Null It Out: Guarding Protected Attributes by Iterative Nullspace Projection](https://mlsanity.com/paper/2004.07667/) (2020)
- [Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classification](https://mlsanity.com/paper/1903.04561/) (2019)
- [Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLP](https://mlsanity.com/paper/2103.00453/) (2021)
- [Large Language Model Safety: A Holistic Survey](https://mlsanity.com/paper/2412.17686/) (2024)
- [Decoding Compressed Trust: Scrutinizing the Trustworthiness of Efficient LLMs Under Compression](https://mlsanity.com/paper/2403.15447/) (2024)
- [MBBQ: A Dataset for Cross-Lingual Comparison of Stereotypes in Generative LLMs](https://mlsanity.com/paper/2406.07243/) (2024)
- [Understanding the Effect of Model Compression on Social Bias in Large Language Models](https://mlsanity.com/paper/2312.05662/) (2023)
- [Investigating the Impact of Quantization Methods on the Safety and Reliability of Large Language Models](https://mlsanity.com/paper/2502.15799/) (2025)
- [Fine-Tuning, Quantization, and LLMs: Navigating Unintended Outcomes](https://mlsanity.com/paper/2404.04392/) (2024)
- [Fairness Definitions in Language Models Explained](https://mlsanity.com/paper/2407.18454/) (2024)
- [The Impact of Inference Acceleration on Bias of LLMs](https://mlsanity.com/paper/2410.22118/) (2024)
- [HarmLevelBench: Evaluating Harm-Level Compliance and the Impact of Quantization on Model Alignment](https://mlsanity.com/paper/2411.06835/) (2024)
- [Mitigating Biases in Language Models via Bias Unlearning](https://mlsanity.com/paper/2509.25673/) (2025)

## Built on by (most cited first, 1 of 1)

- [QuantiBias: Benchmarking Quantization-Induced Bias in LLMs](https://mlsanity.com/paper/2607.21063/) (2026)

## Abstract (arXiv)

Quantization is widely adopted to reduce the computational cost of large language models (LLMs); however, its implications for fairness and safety, particularly in dynamic quantization and multilingual contexts, remain underexplored. In this work, we conduct a systematic study of how static and dynamic quantization methods impact fairness and safety across benchmarks measuring intrinsic and extrinsic bias and safety alignment. For fairness, we evaluate English, French, Dutch, Spanish, and Turkish; for safety, we focus on English, Korean, and Arabic. Our findings reveal that quantization consistently degrades fairness and safety, with dynamic methods demonstrating greater stability than static ones. Moreover, fairness degradation varies across languages, while safety deterioration is especially pronounced in non-English settings. To address these risks, we introduce Critical Weight Protection, a novel technique that identifies and preserves fairness- and safety-critical weights during quantization. This approach effectively mitigates bias and safety deterioration without costly retraining or alignment, maintaining trustworthiness while retaining efficiency.