How Much Do Circuits Tell Us? Measuring the Consistency and Specificity of Language Model Circuits

TL;DR

This paper investigates the properties of circuits in language models, revealing high reuse across tasks but limited task-specificity, which questions their utility for targeted interpretability.

Contribution

It introduces measures of circuit consistency and specificity, and demonstrates that circuits are largely shared across tasks, challenging assumptions about their task-specificity.

Findings

High within-task circuit reuse observed.

Shared circuits are necessary for task performance.

Circuits are not highly task-specific, overlapping across tasks.

Abstract

The circuits framework in mechanistic interpretability aims to identify causally important sparse subgraphs of model components, typically evaluated by measuring necessity and sufficiency. We measure circuit reuse, the proportion of components shared across per-example circuits within a task, and investigate two less-studied properties of this: consistency, the recurrence of components within a task, and specificity, their uniqueness to a task. Using edge attribution patching across six tasks and seven models, we find that within-task reuse is high and that shared components are necessary for task performance, with ablations causing up to $\sim$100% relative accuracy drops. However, circuits turn out not to be task-specific: ablating one task's circuit damages another task's performance about as much as that task's own circuit does. We discover that this is due to substantial overlap between circuits across tasks, which are causally important for performance. Some circuits do contain a smaller set of task-specific components, but these account for only a modest portion of circuit performance. Overall, our findings suggest that while circuit discovery at the level of attention heads and MLP layers identifies important components, their lack of task-specificity raises questions about the degree to which circuits can support targeted understanding and intervention on model behavior.