# Lightweight Safety Guardrails Using Fine-tuned BERT Embeddings

arXiv: https://arxiv.org/abs/2411.14398 · HTML: https://mlsanity.com/paper/2411.14398/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Safety alignment, refusal and guardrails, one of 12 directions in LLM safety and security.
- Builds on 7 of its 12 references that are in the map.
- First posted on arXiv 21 Nov 2024.

## Placed in

- LLM safety and security / [Safety alignment, refusal and guardrails](https://mlsanity.com/safety/safety-alignment-refusal-and-guardrails/)

## Builds on (most cited first, 7 of 7)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations](https://mlsanity.com/paper/2312.06674/) (2023)
- [WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs](https://mlsanity.com/paper/2406.18495/) (2024)
- [NeMo Guardrails: A Toolkit for Controllable and Safe LLM Applications with Programmable Rails](https://mlsanity.com/paper/2310.10501/) (2023)
- [A New Generation of Perspective API: Efficient Multilingual Character-level Transformers](https://mlsanity.com/paper/2202.11176/) (2022)
- [AEGIS: Online Adaptive AI Content Safety Moderation with Ensemble of LLM Experts](https://mlsanity.com/paper/2404.05993/) (2024)

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

- [GuardReasoner: Towards Reasoning-based LLM Safeguards](https://mlsanity.com/paper/2501.18492/) (2025)
- [Detecting and Filtering Unsafe Training Data via Data Attribution with Denoised Representation](https://mlsanity.com/paper/2502.11411/) (2025)
- [Toxicity in Online Platforms and AI Systems: A Survey of Needs, Challenges, Mitigations, and Future Directions](https://mlsanity.com/paper/2509.25539/) (2025)
- [MoGU V2: Toward a Higher Pareto Frontier Between Model Usability and Security](https://mlsanity.com/paper/2509.06807/) (2025)
- [Paved with True Intents: Intent-Aware Training Improves LLM Safety Classification Across Training Regimes](https://mlsanity.com/paper/2606.27210/) (2026)
- [Safety Beyond the Interface: Detecting Harm via Latent States in Large Language Models](https://mlsanity.com/paper/2609.19472/) (2026)
- [Do Encoders Suffice? A Systematic Comparison of Encoder and Decoder Safety Judges for LLM Adversarial Evaluation](https://mlsanity.com/paper/2606.25782/) (2026)
- [Overflip: Repetition-Induced Label Flips in Guardrail Models](https://mlsanity.com/paper/2609.15013/) (2026)

## Abstract (arXiv)

With the recent proliferation of large language models (LLMs), enterprises have been able to rapidly develop proof-of-concepts and prototypes. As a result, there is a growing need to implement robust guardrails that monitor, quantize and control an LLM's behavior, ensuring that the use is reliable, safe, accurate and also aligned with the users' expectations. Previous approaches for filtering out inappropriate user prompts or system outputs, such as LlamaGuard and OpenAI's MOD API, have achieved significant success by fine-tuning existing LLMs. However, using fine-tuned LLMs as guardrails introduces increased latency and higher maintenance costs, which may not be practical or scalable for cost-efficient deployments. We take a different approach, focusing on fine-tuning a lightweight architecture: Sentence-BERT. This method reduces the model size from LlamaGuard's 7 billion parameters to approximately 67 million, while maintaining comparable performance on the AEGIS safety benchmark.