com-jepa 0.1.0

First public research release of Organisational Commitment JEPA.

• Add a pinned TabPFN-3 / PyTorch GPU comparison using identical eligible task data, individual test-row predictions, checkpoint verification, and provenance reports. Record the first synthetic GB10 run in the comparison guide. Model weights are separately licensed and are not distributed.
• Add an optional scikit-learn Random Forest demo on explicitly invented tabular snapshots, with a training-only base-rate comparator, temporal label eligibility, exclusion accounting, reproducible reports, and ML CI. No real organisational findings are implied.
• Add eight fictional challenge cards, a contract-question map, a case template, and a partner discovery guide. Keep their unresolved semantics outside the unchanged 0.1.0 executable contract.
• Establish the organisational thesis, claims ledger, and baseline-first research plan.
• Propose a draft 0.1.0 commitment event export with version lineage and occurrence/availability timestamps.
• Add a wholly fictional trajectory, context projection utility, and temporal/semantic validation tests.
• Describe application responsibilities, pilot work packages, and a Birds of a Feather session.
• Establish contribution, attribution, governance, conduct, citation, licensing, and security guidance.

Validation: 25 unit tests pass with the ML dependencies installed; the fictional event contract and historical snapshot checks pass. The pinned TabPFN checkpoint was also executed on a GPU. The synthetic result is an engineering demonstration and does not establish organisational validity.

No real organisational dataset, model weights, validated organisational predictor, JEPA implementation, or production integration is distributed in this release. The event contract remains a draft and may change incompatibly during 0.x releases.