# Biases in Edge Language Models: Detection, Analysis, and Mitigation

arXiv: https://arxiv.org/abs/2502.11349 · HTML: https://mlsanity.com/paper/2502.11349/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 45 references that are in the map.
- First posted on arXiv 17 Feb 2025.

## Builds on (most cited first, 17 of 17)

- [Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context](https://mlsanity.com/paper/2403.05530/) (2024)
- [Pruning Filters for Efficient ConvNets](https://mlsanity.com/paper/1608.08710/) (2016)
- [Justice or Prejudice? Quantifying Biases in LLM-as-a-Judge](https://mlsanity.com/paper/2410.02736/) (2024)
- [Fine-Tuning Language Models with Just Forward Passes](https://mlsanity.com/paper/2305.17333/) (2023)
- [Break the Sequential Dependency of LLM Inference Using Lookahead Decoding](https://mlsanity.com/paper/2402.02057/) (2024)
- [Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scale](https://mlsanity.com/paper/2211.03759/) (2022)
- ["Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letters](https://mlsanity.com/paper/2310.09219/) (2023)
- [LaMini-LM: A Diverse Herd of Distilled Models from Large-Scale Instructions](https://mlsanity.com/paper/2304.14402/) (2023)
- [Is ChatGPT Fair for Recommendation? Evaluating Fairness in Large Language Model Recommendation](https://mlsanity.com/paper/2305.07609/) (2023)
- [A Review on Edge Large Language Models: Design, Execution, and Applications](https://mlsanity.com/paper/2410.11845/) (2024)
- [Systematic Biases in LLM Simulations of Debates](https://mlsanity.com/paper/2402.04049/) (2024)
- [On Protecting the Data Privacy of Large Language Models (LLMs): A Survey](https://mlsanity.com/paper/2403.05156/) (2024)
- [Optimize Weight Rounding via Signed Gradient Descent for the Quantization of LLMs](https://mlsanity.com/paper/2309.05516/) (2023)
- [CFaiRLLM: Consumer Fairness Evaluation in Large-Language Model Recommender System](https://mlsanity.com/paper/2403.05668/) (2024)
- [Empirical Guidelines for Deploying LLMs onto Resource-constrained Edge Devices](https://mlsanity.com/paper/2406.03777/) (2024)
- [Designing Efficient LLM Accelerators for Edge Devices](https://mlsanity.com/paper/2408.00462/) (2024)
- [Addressing the Memory Bottleneck in AI Model Training](https://mlsanity.com/paper/2003.08732/) (2020)

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

- [Unified Threat Detection and Mitigation Framework (UTDMF): Combating Prompt Injection, Deception, and Bias in Enterprise-Scale Transformers](https://mlsanity.com/paper/2510.04528/) (2025)

## Abstract (arXiv)

The integration of large language models (LLMs) on low-power edge devices such as Raspberry Pi, known as edge language models (ELMs), has introduced opportunities for more personalized, secure, and low-latency language intelligence that is accessible to all. However, the resource constraints inherent in edge devices and the lack of robust ethical safeguards in language models raise significant concerns about fairness, accountability, and transparency in model output generation. This paper conducts a comparative analysis of text-based bias across language model deployments on edge, cloud, and desktop environments, aiming to evaluate how deployment settings influence model fairness. Specifically, we examined an optimized Llama-2 model running on a Raspberry Pi 4; GPT 4o-mini, Gemini-1.5-flash, and Grok-beta models running on cloud servers; and Gemma2 and Mistral models running on a MacOS desktop machine. Our results demonstrate that Llama-2 running on Raspberry Pi 4 is 43.23% and 21.89% more prone to showing bias over time compared to models running on the desktop and cloud-based environments. We also propose the implementation of a feedback loop, a mechanism that iteratively adjusts model behavior based on previous outputs, where predefined constraint weights are applied layer-by-layer during inference, allowing the model to correct bias patterns, resulting in 79.28% reduction in model bias.