大语言模型如何遵循指令：技能协调，而非通用机制
How LLMs Follow Instructions: Skillful Coordination, Not a Universal Mechanism
AI总结：
研究通过九种不同任务分析发现，大语言模型遵循指令并非依赖通用机制，而是通过协调多种语言能力实现，且存在任务间转移弱、依赖稀疏等特点。
摘要：
指令微调通常被认为赋予语言模型跨领域遵循指令的能力，但其底层机制仍不明确。遵循指令是依赖通用机制还是组合技能部署？我们通过在三个指令微调模型上进行九种不同任务的诊断探测来探讨这一问题。我们的分析提供了反对通用机制的证据。首先，跨任务训练的一般探测器在所有任务上均表现不佳，表明表示共享有限。其次，跨任务迁移弱且按技能相似性聚类。第三，因果消融揭示了稀疏的不对称依赖而非共享表示。任务还按复杂性在层间分层，结构性约束早期出现，语义任务后期出现。最后，时间分析显示约束满足在生成过程中作为动态监控，而非生成前的规划。这些发现表明，遵循指令更好地表现为多种语言能力的协调，而非单一抽象约束检查过程的部署。
英文摘要：
Instruction tuning is commonly assumed to endow language models with a domain-general ability to follow instructions, yet the underlying mechanism remains poorly understood. Does instruction-following rely on a universal mechanism or compositional skill deployment? We investigate this through diagnostic probing across nine diverse tasks in three instruction-tuned models. Our analysis provides converging evidence against a universal mechanism. First, general probes trained across all tasks show selective rather than uniform deficits relative to task-specific specialists, indicating that representational sharing is partial and structured rather than global. Second, cross-task transfer is weak and clustered by skill similarity. Third, causal ablation reveals sparse asymmetric dependencies rather than shared representations. Tasks also stratify by complexity across layers, with structural constraints emerging early and semantic tasks emerging late. Finally, temporal analysis shows that the constraint signal becomes decodable only once generation is under way, and remains so throughout the response. These findings indicate that instruction-following is better characterized as skillful coordination of diverse linguistic capabilities rather than deployment of a single abstract constraint-checking process.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
