Results and documents
Latest rating run (4): no improvement over the context model under the pre-registered rule; no ordered ranking is given. Report · commander-ratings-v4.json
- Overview
README.md - Rating run 4: memo
docs/research/commander-ratings-v4.md - Rating run 4: report
artifacts/commander-ratings-v4.md - Rating run 4: design
docs/commander-ratings-v4.md - Rating run 3: memo
docs/research/commander-ratings-v3.md - Rating run 3: report
artifacts/commander-ratings-v3.md - Rating run 3: how sure is the verdict?
artifacts/commander-ratings-v3-uncertainty.md - Extraction audit
docs/extraction-audit.md - Methodology
docs/methodology.md - Evidence contract
docs/evidence-contract.md - Roadmap
docs/roadmap.md