Capability Conditioned Scaffolding for Professional Human LLM Collaboration

面向专业人类与大模型协作的“能力条件化脚手架”

Abstract: Large language model personalization typically adapts outputs to user preferences and style but does not account for differences in user evaluation capacity across domains of expertise. This limitation can encourage Professional Domain Drift, where users rely on AI generated reasoning in domains they cannot reliably evaluate.

摘要： 大语言模型的个性化通常侧重于根据用户的偏好和风格调整输出，却忽略了用户在不同专业领域中评估能力的差异。这种局限性可能导致“专业领域漂移”（Professional Domain Drift），即用户在自身无法可靠评估的领域中过度依赖 AI 生成的推理结果。

We introduce Capability Conditioned Scaffolding, a typed framework that partitions expertise into strong, mixed, and weak domains and conditions intervention behavior on structured capability profiles.

我们引入了“能力条件化脚手架”（Capability Conditioned Scaffolding），这是一个类型化框架，它将专业知识划分为强、中、弱三个领域，并根据结构化的能力画像来调节 AI 的干预行为。

A pilot evaluation across multiple MMLU subsets and four LLM substrates shows consistent profile conditioned intervention behavior, including categorical inversion under profile swapping and selective activation in mixed domain risk zones.

通过对多个 MMLU 子集和四种大模型底座进行的初步评估显示，该框架展现了稳定的一致性：在基于画像的干预行为中，当切换用户画像时，AI 的干预策略会发生类别反转，并在混合领域的风险区域表现出选择性激活。

These findings suggest that capability aware scaffolding can support more reliable professional human AI collaboration beyond stylistic personalization.

这些研究结果表明，具备能力感知功能的脚手架不仅能实现风格化的个性化，还能支持更可靠的专业级人机协作。