MASteer: Multi-Agent Adaptive Steer Strategy for End-to-End LLM Trustworthiness Repair

TL;DR

MASteer introduces an end-to-end framework utilizing multi-agent systems and adaptive representation engineering to improve trustworthiness in large language models efficiently and automatically.

Contribution

It is the first framework to automate trustworthiness repair in LLMs using adaptive, context-aware representation strategies with multi-agent sample generation.

Findings

Outperforms baselines on trustworthiness metrics.

Improves LLaMA-3.1-8B-Chat by 15.36%.

Enhances Qwen-3-8B-Chat by 4.21%.

Abstract

Large Language Models (LLMs) face persistent and evolving trustworthiness issues, motivating developers to seek automated and flexible repair methods that enable convenient deployment across diverse scenarios. Existing repair methods like supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) are costly and slow, while prompt engineering lacks robustness and scalability. Representation engineering, which steers model behavior by injecting targeted concept vectors during inference, offers a lightweight, training-free alternative. However, current approaches depend on manually crafted samples and fixed steering strategies, limiting automation and adaptability. To overcome these challenges, we propose MASteer, the first end-to-end framework for trustworthiness repair in LLMs based on representation engineering. MASteer integrates two core components: AutoTester, a multi-agent system that generates diverse, high-quality steer samples tailored to developer needs; and AutoRepairer, which constructs adaptive steering strategies with anchor vectors for automated, context-aware strategy selection during inference. Experiments on standard and customized trustworthiness tasks show MASteer consistently outperforms baselines, improving metrics by 15.36% on LLaMA-3.1-8B-Chat and 4.21% on Qwen-3-8B-Chat, while maintaining general model capabilities. MASteer demonstrates strong robustness, generalization, and practical value for scalable, efficient trustworthiness repair.