domains / families / research / pyvene

pyvene

separate

Provides guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention framework. Use when conducting causal tracing, activation patching, interchange intervention training, or testing causal hypotheses about model behavior.

When to use

• Perform causal tracing (ROME-style localization)
• Run activation patching experiments
• Conduct interchange intervention training (IIT)
• Test causal hypotheses about model components
• Share/reproduce intervention experiments via HuggingFace
• Work with any PyTorch architecture (not just transformers)
• You need exploratory activation analysis → Use **TransformerLens**
• You want to train/analyze SAEs → Use **SAELens**

When not to use

• exploratory activation analysis
• training or analyzing SAEs
• remote execution on massive models
• lower-level control

Skill page

Source