# S3LoRA: Safe Spectral Sharpness-Guided Pruning in Adaptation of Agent Planner

arXiv: https://arxiv.org/abs/2508.15068 · HTML: https://mlsanity.com/paper/2508.15068/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use.
- Builds on 26 of its 44 references that are in the map.
- First posted on arXiv 20 Aug 2025.

## Placed in

- LLM agents and tool use / [Agent threats, guardrails and governance](https://mlsanity.com/agents/agent-threats-guardrails-and-governance/)

## Builds on (most cited first, 26 of 26)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017)
- [Sparks of Artificial General Intelligence: Early experiments with GPT-4](https://mlsanity.com/paper/2303.12712/) (2023)
- [A Survey on Large Language Model based Autonomous Agents](https://mlsanity.com/paper/2308.11432/) (2023)
- [The Rise and Potential of Large Language Model Based Agents: A Survey](https://mlsanity.com/paper/2309.07864/) (2023)
- [LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day](https://mlsanity.com/paper/2306.00890/) (2023)
- [Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!](https://mlsanity.com/paper/2310.03693/) (2023)
- [SAMSum Corpus: A Human-annotated Dialogue Dataset for Abstractive Summarization](https://mlsanity.com/paper/1911.12237/) (2019)
- [AgentTuning: Enabling Generalized Agent Abilities for LLMs](https://mlsanity.com/paper/2310.12823/) (2023)
- [Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models](https://mlsanity.com/paper/2310.02949/) (2023)
- [Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications](https://mlsanity.com/paper/2402.05162/) (2024)
- [Survey on Evaluation of LLM-based Agents](https://mlsanity.com/paper/2503.16416/) (2025)
- [Spectral Norm Regularization for Improving the Generalizability of Deep Learning](https://mlsanity.com/paper/1705.10941/) (2017)
- [Removing RLHF Protections in GPT-4 via Fine-Tuning](https://mlsanity.com/paper/2311.05553/) (2023)
- [OpenAgents: An Open Platform for Language Agents in the Wild](https://mlsanity.com/paper/2310.10634/) (2023)
- [AgentGen: Enhancing Planning Abilities for Large Language Model based Agent via Environment and Task Generation](https://mlsanity.com/paper/2408.00764/) (2024)
- [SVD-LLM V2: Optimizing Singular Value Truncation for Large Language Model Compression](https://mlsanity.com/paper/2503.12340/) (2025)
- [AgentBank: Towards Generalized LLM Agents via Fine-Tuning on 50000+ Interaction Trajectories](https://mlsanity.com/paper/2410.07706/) (2024)
- [Approaching Deep Learning through the Spectral Dynamics of Weights](https://mlsanity.com/paper/2408.11804/) (2024)
- [ATLaS: Agent Tuning via Learning Critical Steps](https://mlsanity.com/paper/2503.02197/) (2025)
- [A Framework for Benchmarking and Aligning Task-Planning Safety in LLM-Based Embodied Agents](https://mlsanity.com/paper/2504.14650/) (2025)
- [Safe Pruning LoRA: Robust Distance-Guided Pruning for Safety Alignment in Adaptation of LLMs](https://mlsanity.com/paper/2506.18931/) (2025)
- [Matrices with Gaussian noise: optimal estimates for singular subspace perturbation](https://mlsanity.com/paper/1803.00679/) (2018)
- [Robust SVD Made Easy: A fast and reliable algorithm for large-scale data analysis](https://mlsanity.com/paper/2402.09754/) (2024)
- [On the Interaction of Compressibility and Adversarial Robustness](https://mlsanity.com/paper/2507.17725/) (2025)

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

- [Learn the Directions, Normalize the Gains: Post-Training Normalization for LoRA](https://mlsanity.com/paper/2610.02067/) (2026)

## Abstract (arXiv)

Adapting Large Language Models (LLMs) using parameter-efficient fine-tuning (PEFT) techniques such as LoRA has enabled powerful capabilities in LLM-based agents. However, these adaptations can unintentionally compromise safety alignment, leading to unsafe or unstable behaviors, particularly in agent planning tasks. Existing safety-aware adaptation methods often require access to both base and instruction-tuned model checkpoints, which are frequently unavailable in practice, limiting their applicability. We propose S3LoRA (Safe Spectral Sharpness-Guided Pruning LoRA), a lightweight, data-free, and model-independent framework that mitigates safety risks in LoRA-adapted models by inspecting only the fine-tuned weight updates. We first introduce Magnitude-Aware Spherically Normalized SVD (MAS-SVD), which robustly analyzes the structural properties of LoRA updates while preserving global magnitude information. We then design the Spectral Sharpness Index (SSI), a sharpness-aware metric to detect layers with highly concentrated and potentially unsafe updates. These layers are pruned post-hoc to reduce risk without sacrificing task performance. Extensive experiments and ablation studies across agent planning and language generation tasks show that S3LoRA consistently improves safety metrics while maintaining or improving utility metrics and significantly reducing inference cost. These results establish S3LoRA as a practical and scalable solution for safely deploying LLM-based agents in real-world, resource-constrained, and safety-critical environments.