domains / families / research / nnsight

nnsight

separate

Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution. Use when needing to run interpretability experiments on massive models (70B+) without local GPU resources, or when working with any PyTorch architecture.

When to use

• Run interpretability experiments on models too large for local GPUs (70B, 405B)
• Work with any PyTorch architecture (transformers, Mamba, custom models)
• Perform multi-token generation interventions
• Share activations between different prompts
• Access full model internals without reimplementation
• You want consistent API across models → Use **TransformerLens**
• You need declarative, shareable interventions → Use **pyvene**
• You're training SAEs → Use **SAELens**

When not to use

• Consistent API across models
• Declarative, shareable interventions
• Training SAEs
• Working only with small models locally

Skill page

Source