mechanistic-interpretability-nnsight

原文涵盖的主题

• nnsight: Transparent Access to Neural Network Internals
• Key Value Proposition
• When to Use nnsight
• Installation
• Core Concepts
• LanguageModel Wrapper
• Tracing Context
• Proxy Objects

nnsight (/ɛn.saɪt/) enables researchers to interpret and manipulate the internals of any PyTorch model, with the unique capability of running the same code locally on small models or remotely on massive models (70B+) via NDIF. GitHub: ndif-team/nnsight (730+ stars) Paper: NNsight and NDIF: Democratizing Access to Foundation Model Internals (ICLR 2025) Write once, run anywhere: The same interpretability code works on GPT-2 locally or Llama-3.1-405B remotely. Just toggle remote=True. Use nnsight when you need to: 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 Consider alternatives when: You want consistent API across models → Use TransformerLens You need declarative, shareable interventions → Use pyvene You're training SAEs → Use SAELens You only work with small models locally → TransformerLens may be simpler For remote NDIF execution, sign up at login.ndif.us for an API key. The trace context manager enables deferred execution - operations

这里只展示节选，请以完整原文和当前要求为准。