Diagnose, Localize, Align: A Full-Stack Framework for Reliable LLM Multi-Agent Systems under Instruction Conflicts

Guancheng Wan, Leixin Sun, Longxu Dou, Zitong Shi, Fang Wu, Eric Hanchen Jiang, Wenke Huang, Guibin Zhang, Hejia Geng, Xiangru Tang, Zhenfei Yin, Yizhou Sun, Wei Wang

Keywords: LLM, LLM Agent

Abstract: Large Language Model (LLM)-powered multi-agent systems (MAS) have rapidly advanced collaborative reasoning, tool use, and role-specialized coordination in complex tasks. However, reliability-critical deployment remains hindered by a systemic failure mode: **hierarchical compliance** under **instruction conflicts** (system–user, peer–peer), where agents misprioritize system-level rules in the presence of competing demands. Moreover, widely used macro-level metrics (e.g., pass@k) obscure these micro-level violations and offer little actionable guidance for remedy. In this work, we present a full-stack, three-stage framework: (1) **Diagnose** - *Contextualized Role Adherence Score* (CRAS), a query-wise, context-aware scoring metric that decomposes role adherence into four measurable dimensions; (2) **Localize** - attention drift analysis revealing that instruction conflicts are resolved by attention heads that are largely concentrated in middle layers; (3) **Align** - *Surgical Alignment of Instruction Layers (SAIL)*, which installs LoRA only on the localized focal layers and optimizes a token-weighted DPO-style preference objective that credits tokens by their focal attentional contribution. Across standard benchmarks and MAS frameworks, our surgical approach improves instruction hierarchy compliance (e.g., +5.60% with AutoGen on MedQA) without full-model finetuning. The code is available at [https://anonymous.4open.science/r/DLA-ICLR-6DF6/](https://anonymous.4open.science/r/DLA-ICLR-6DF6/).

Primary Area: foundation or frontier models, including LLMs

Submission Number: 4364