What problem does it solve? It turns a stable research proposal or method idea into a concrete, paper-oriented experiment plan, so you know exactly which experiments to run, in what order, and what evidence each one must produce to defend your claims. ## Core Features & Use Cases - Claim Mapping: Freezes primary and supporting claims, defines anti-claims to rule out, and links every experiment block to the claim it defends. - Experiment Block Specification: Fully specifies datasets, provenance, baselines, ablations, metrics, success criteria, and failure interpretations for each block. - Run Order and Budgeting: Produces milestone-based execution stages with compute estimates, decision gates, and risk mitigations, separating must-run from nice-to-have runs. - Use Case: After refining a novel LLM interpretability method, use this Skill to generate EXPERIMENT_PLAN.md and EXPERIMENT_TRACKER.md files that define the main table, ablation matrix, and the first three runs to launch. ## Quick Start Ask the assistant to create a detailed experiment plan for your refined research proposal, including ablations, baselines, run order, and compute budget.
