# About these records These documents were written during development and edited for publication: - `preregistration.md`: the rules of every test, committed on 2026-11-03 before any simulation run, with the amendments made later under its section 9; - `validation-log.md`: every attempt, failure, diagnostic and decision, in order, under dated headings; - `results-summary.md`: every result with its value, threshold and verdict; - `connectome-precheck.md`: the structural pre-check of the three chains, made before any simulation; - `results/`: the machine-written reports and the per-trial summaries the tests were judged on. Internal project references (milestone names, commit hashes, CI run IDs, and references to the application the model was built for) were removed or replaced with plain descriptions. Rules, numbers, verdicts and the order of events are unchanged. The reports in `results/` are re-created exactly by the code in `flysim/` from the per-trial summaries next to them. From `flysim/`: ```bash uv run python -m flysim.validation.chains judge --trials ../records/results/chain-tests-trials.json --out chains.md uv run python -m flysim.validation.culling judge --full ../records/results/chain-tests-trials.json \ --variant ../records/results/culling-min{2,3,5,10}-trials.json --out culling.md uv run python -m flysim.validation.speed report --results ../records/results/speed-vec.json --out speed-vec.md uv run python -m flysim.validation.weight_scale report --results ../records/results/weight-scale.json --out weight-scale.md ``` The spike files themselves were never kept. The seeds are fixed, so the `run` commands regenerate them.