TransformerLens Mechanistic Interpretability

This skill empowers researchers and developers to reverse-engineer transformer models using TransformerLens, the industry-standard library for mechanistic interpretability. It provides structured guidance and implementation patterns for complex tasks like activation patching, causal tracing, and circuit analysis. By leveraging HookedTransformer interfaces, users can perform deep-dive investigations into model internals, identify induction heads, and conduct direct logit attribution to understand the underlying algorithms learned during model training.

주요 기능

사용 사례