Use these docs as a map of the repository, not as a substitute for the tests or research artifacts.
| Document | Read it for |
|---|---|
| onboarding.md | Start here to contribute: reading order, the rules every result meets, open problems, and the process constraints that have cost results. |
| api.md | Imports, public objects, compatibility shims, and common commands. |
| architecture.md | Runtime layers, engines, automata helpers, language modules, and archive layout. |
| scientific_status.md | Which claims are tested in the package, which are experimental, and which belong to the literature. |
| supported_surfaces.md | What is maintained as package code, what is compatibility code, and what is research-only. |
| project_context.md | Project history, authorship, collaboration context, and research motivation. |
| references.md | Papers and background literature behind the package and research program. |
| literature.md | Full AC field map, implementation parity matrix, and gap priorities. |
| packaging.md | Release workflow for the neural-assemblies PyPI package. |
| contributing.md | Contributor setup and expected checks. |
| documentation_style.md | Writing standards for docs in this repo. |
| register.md | Every adopted result with its evidence and caveats, rendered from neural_assemblies/theory.py. |
| ../research/plans/PAPERS.md | The papers the register supports, in order of readiness, and what each still needs. |
| ../research/notes/README.md | The reading map for the registrations and design notes: what each line concluded, which file to open first. |
| ../research/experiments/README.md | The active experiment scripts by line, with typical runs; results are in ../research/results/. |
The research tree has its own workflow:
- research/README.md
- research/literature/index.json
- research/claims/index.json
- research/core_questions/index.json
Package docs should not overrule those artifacts. If a scientific statement depends on a particular experiment, link to the experiment or the indexed claim.