transformer-lens (AI-Research-SKILLs)

Agent SkillsFreeOpen Source

Mechanistic-interpretability skill for using TransformerLens to probe LLM internals.

🛡️ AgentReady threat assessment

MAESTRO 7-layer threat model + OWASP AIVSS risk score for transformer-lens (AI-Research-SKILLs), derived from its capabilities.

AIVSS 9.2 · Critical

View MAESTRO 7-layer threat model →

These scores are auto-generated from public information (the agent's own listing, docs, and repository) using the canonical OWASP AIVSS formula and the MAESTRO framework — an estimate for guidance, not a penetration test, audit, or certification. See the scoring methodology — every score is re-derived by the same automated method as an agent's public evidence changes.

Overview

Part of Orchestra-Research's AI research skills library — a skill for mechanistic interpretability work with the TransformerLens library (activation caching, hooks, circuit analysis). Surface: injects library-specific guidance and writes/runs Python interpretability code.

Key features and capabilities

TransformerLens hooks and activation caching
Circuit-analysis workflows
One of a large AI-research skill library

Use cases

Probe an LLM's internal activations
Run a mechanistic-interpretability experiment