Navigate

Quick find

Navigate NeuroForge

Esc

Start typing to search public pages and evidence.

    Reviewed 2026-09-05

    What must be proved next.

    This is not a calendar of promises. Each item names the result that must exist before the programme can move.

    This page shows future decision gates, not delivery dates. For what exists today, use the status page.

    Next decisive tests

    The next result that could change each decision.

    Running an experiment is not enough. A workstream moves only when its stated exit closes.

    01research

    Resolve native transfer and stored-capacity scaling

    The next research question is whether richer cognition and greater stored capability improve useful quality while the active path stays constrained.

    Exit criterion

    Complete qualified fixed-holdout, stored-capacity-matched and active-path-matched comparisons with repeated independent evaluation.

    Current ERAIS research
    02research

    Establish Fracture quality preservation

    Fracture is the migration path for existing neural capability; the decisive question is whether useful task behaviour survives conversion under policy-governed evaluation.

    Exit criterion

    Pass bounded quality, safety and resource gates on qualified source models before any production conversion commitment.

    Governed conversion research
    03governance

    Earn independent partner operation

    Move from founder-run evidence to a bounded partner environment with named owners, reproducible acceptance criteria and an explicit handoff path.

    Exit criterion

    Qualify one partner problem with evidence access, security boundaries, a decision owner and an operational handoff owner.

    Design-partner qualification

    Not yet earned

    These outcomes are not available yet.

    They stay visible, but held. Time, confidence, and effort do not release them; the named evidence gate does.

    governance

    Historical evidence assembly remains held

    This hold describes the July release candidate. It does not date or suppress the newer current systems receipts.

    Release only after

    Regenerate a new evidence release from one authoritative validated handoff before treating it as current publication authority.

    governance

    Open public inference access

    The current chat route is a boundary explanation, not a live ERAIS service.

    Release only after

    Review a bounded inference backend plus security, privacy, monitoring, rollback and capacity controls.

    commercial

    Production commitments

    Research execution and bounded professional engagements are not equivalent to a production ERAIS product.

    Release only after

    Establish repeatable quality, security, observability, recovery and operating-cost envelopes in an independently run environment.

    research

    Broad quality or efficiency superiority

    Mechanism evidence and source-pinned receipts remain public with their scope; promotion waits for fair, repeated, matched-quality comparisons.

    Release only after

    Reconcile qualified dense controls, fixed holdouts and full resource accounting under an approved claim review.

    How a result moves

    Prove what ran. Say what it means. Earn the right to offer it.

    01 / Source

    Prove what ran

    Bind the result to an exact source, approved artifacts, digests, and final verification—not a moving worktree.

    02 / Meaning

    Say what it means

    State what was observed, what it supports, what it does not support, and what evidence would change the decision.

    03 / Availability

    Earn the right to offer it

    A technical result does not become a product. Operations, rights, security, recovery, and delivery must pass separately.

    Go deeper

    Inspect the exact claim or bring us one research question.