pyvene-interventions
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.
- Score
- Unranked
- Evidence
- No signals yet
- Last commit
- Not known
- Listed
Install
armory install pyvene-interventions --cli claudewrites the skill into.claude/skills/pyvene-interventions/SKILL.mdListed as compatible
# fetches the source and writes it to:
.claude/skills/pyvene-interventions/SKILL.mdNeeds the armory CLI · not on npm yet, build it from cli/ in the repository
What it is
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 it
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.
How to install / invoke
# Copy the skill into your .claude/skills/ directory
curl -sL https://raw.githubusercontent.com/davila7/claude-code-templates/main/cli-tool/components/skills/ai-research/mechanistic-interpretability-pyvene/SKILL.md -o .claude/skills/mechanistic-interpretability-pyvene/SKILL.md
Notes
Extracted from davila7/claude-code-templates, mechanistic-interpretability-pyvene category.