What problem does it solve?
This skill helps you design and present evaluation experiments so your robotics and embodied-AI paper’s claims are supported by measurable, fair, and reproducible results.
Core Features & Use Cases
- Claim-to-experiment mapping: translate each claimed module or contribution into at least one direct experiment or diagnostic rather than relying on overall system performance.
- Metric protocol coverage: choose appropriate metric families for VLN/navigation, SLAM/state estimation, and control/RL, including APE/RPE, SPL, success and collision rates, runtime, and ablations.
- Fairness and statistical reporting: ensure consistent splits, modality parity, budget parity, correct handling of failures/crashes/timeouts, and clear guidance on seeds, confidence intervals, and significance wording.
- Paper-ready reporting rules: produce LaTeX-ready tables/figures guidance with explicit measurement scope, episode counts, and simulation-vs-hardware labeling.
Quick Start
Use this skill to generate a complete experiment-and-metrics plan for a robotics paper by providing your paper’s claims, target tasks (e.g., VLN/SLAM/control), and available hardware/simulator setup, then ask it to output an evaluation design that includes baselines, metrics, ablations, run counts, and reporting rules.
