Signals Programme — Relational Domain Adaptation (aegir workstream)

Status: programme charter (2026-06-16; living — updated 2026-07-09: tri-root lineup, inverted topic layer, catalog-is-derived). The go-forward development programme for the aegir track, framed as a workstream of the cross-repo Signals initiative. Builds on end_to_end_and_meta_harness.md (the substrate-evolution machinery) and the repo-root EVIDENCE.md gate discipline. Designed to be lifted into a GitHub project across the component repos (§Component map).

Thesis

Signals co-develops a bounded, verifiable, signal-driven agent ecosystem (Holland Signals & Boundaries; intrinsic verifiability per Gaius RASE). This workstream is its relational domain-adaptation engine, and it rests on three commitments established over the inc-2 work:

1. The ontology is the domain-adaptive surface — not weight-bound software. A de-novo-curated, reasoner-verified ontology carries domain meaning independent of any model, so it can be evolved before a trusted model exists (the regime where a meta-harness cannot operate).
2. The reasoner (HermiT) is the model-independent oracle — sound & complete, so coherence and realization are ground truth, not learned proxies. It certifies formal correctness; the corpus and the model certify domain correctness. All three anchors stay live (anti-folie-à-deux).
3. The H-Net+RWKV model is the ultimate fitness measure — trained from scratch on the ontology-grounded corpus, evaluated on relational understanding and de novo Data Element elucidation with conventional evals/post-training, deliberately kept as a trustworthy, non-co-adapting arbiter.

The generator (SAE-instrumented Qwen, fine-tuned to mint the ontology) and the downstream model (H-Net+RWKV, from scratch) are distinct by design: experiment upstream, keep the instrument boring.

Source of truth & data flow — the ontology is primary; everything else serves it

The single source of truth is the generated ontology artifact itself. Atlas, Qdrant, the tri-root lineup projection (build/dev/{current,scratch,archive}), and even the published sdg-corpora are projections, indices, views, and exports of it — never competing stores. The ontology spans two disclosure tiers: KNOW (the full, curated working ontology) ⊇ SHARE (sdg-corpora, the published subset). Same artifact, two tiers — not two artifacts.

• Dependency arrows point inward. Regenerating the ontology re-projects the lineup trunk (just kb-build), re-syncs Atlas, re-indexes Qdrant, re-exports sdg-corpora. That inward-pointing dependency is what keeps a multi-store assembly coherent instead of a web of drifting masters.
• The lineup ⟷ Atlas stay in sync because both project from the ontology — not via a direct link. There is no lineup↔Atlas channel; both are downstream of the one SoT.
• Atlas edits suggest, they do not commit. Atlas is a rich glossary-editing surface, but a curator’s edit there is a proposal that round-trips into the ontology’s curation queue (reviewed, reasoner-gated, applied), then re-projects outward. So the Atlas glossary-sync is a suggestion-returning projection, not an authoritative store — its write path is a PR against the ontology, never a commit to it. (Same shape as trunk → release promotion; Atlas is just another suggestion inbox alongside the generator’s minted candidates and the authored scratch notes.)
• The lineup is the navigate / explore / curate layer — the one place ontology-projection, Atlas-sync, and the Qdrant retrieval text are seen together, and from which curation decisions are made. Its three roots are refs (RH 2026-07-06): current = the latest sdg-corpora release as a complete kasten (released corpus + its era lexicon + released DDL + card); scratch = trunk (the full live projection — live catalog, DDL spine, generated web, accreting corpus, zettel chain, strategy, items); archive = past releases + frozen snapshots. Promotion = snapshot-freeze current → archive, project the new release, trunk rolls on; the same surface ids resolve per root (gateway ?root=).

The verification membrane (corollary of commitment #2). The anchors partition verification by where reality is authored. Inside the loop — the ontology and everything projected from it (DDL spine, corpus structure) — verification is intrinsic: HermiT is the oracle, correctness is proved, not checked. Validation-after-the-fact (shape checkers, assertion contracts, external catalogs) compensates for not generating-correct-by-construction and has no place inside the boundary. Extrinsic verification is legitimate only at the membrane where un-generated reality enters — grounding the corpus against FinePDFs (R1) and the held-out downstream eval — where the system honestly tests itself against a world it did not author. So anchors #2 (formal / intrinsic) and #3 (domain / membrane) are both required and not redundant: they verify opposite sides of the boundary. Consequence — integrate freely, depend on nothing. Under in-situ RASE engineering, switching costs across the metadata/governance plane (OpenMetadata, OpenLineage, Atlas, Ranger, …) are ~zero, so integration carries no lock-in — which licenses deep, enthusiastic integration rather than abstention. We implement the standing directives in full: OL/Marquez compatibility (external tooling sees our lineage events) and Atlas deep integration (a richly extended glossary/lineage surface). This is non-dependence, not non-integration. The discipline that keeps it non-dependent: the ontology is the single source of truth, so Atlas and every projection stay rebuildable-from-the-ontology-or-it’s-a-bug, never a master (Atlas edits suggest, never commit). We sample patterns and adapt protocols on our terms; we owe the plane no gravity — and that freedom is exactly what makes integrating with it generously safe.

The maturation arc — substantially landed. The seed-crystal stage is over: the hand-authored seed families are retired and the catalog is fully derived (src/aegir/ontology/catalog/catalog.json, accreted by the membrane-gated derive → promote loop; discovery via schema.catalog_files()). The terms-are-real end state arrived as the inverted topic layer (2026-07-07, phase-gated PASS — the gate record is authoritative): topics ≡ ontology-grounded concept anchors in Qdrant (sdg_topics: the 29 SKOS domains + all 433 live catalog terms; src/aegir/ontology/topic_layer.py), items = anchor-proportional passage windows, one item → one topic via hierarchical-margin MaxSim or UNASSIGNED — and the lexicon is the parameter: unaligned input indicts the topics, never the input (the inverted-LDA direction). Every association is content-addressed to the collection state that adjudicated it, so accumulated lineage survives registry evolution. The corpus-fitted BERTopic-era topic model (topic_alignment.py, build_topic_model.py, T_I.pkl) is deprecated — v0.3 reproducibility only; R_D’s successor is congruence (src/aegir/ontology/congruence.py) + the topic layer. The end state stands and is now measurable: the lineup ≡ Atlas glossary ≡ Qdrant index — three views of one real ontology.

The two faces that make a term “real”: (1) Qdrant augmentation, recorded as SKOS annotation properties (built) — per the BERTSubs §4.3.2 multi-label technique, each term’s distinguishing text-features live in the ontology as skos:prefLabel / skos:altLabel (the BERTSubs multi-label set / MaxSim match surfaces) / skos:definition / skos:scopeNote / skos:example — common SKOS constructs with domain values, not novel ones. The lineup panel renders them, they assemble into the ColBERT/MaxSim retrieval text, and L(c1)×L(c2) over the altLabel sets multiplies the BERTSubs subsumption pairs (→ the hierarchy edges for per-term-panel navigation). The surfaces are now authored through a closed membrane loop (scripts/author_skos_surfaces.py): the engine proposes {alt_labels, scope_note, definition} per term; the annotation membrane (src/aegir/ontology/annotation_membrane.py, gates M1–M6 + M8: positive voice only — definition-by-negation is an embedding anti-pattern) disposes with a reason; the result is verified by the self-retrieval margin gate plus the input-side M7 basin gate — the annotation-layer oracle, distinct from HermiT on the axiom layer (annotations never assert taxonomy). The subsumption hierarchy over those terms is now realized autonomously by mediate_hierarchy (scripts/, built 2026-06-17): mpnet candidates → Grok proposes (ACP) → a two-layer gate — HermiT consistency/coherence/acyclicity and domain vocabulary overlap — admits only verified edges, no human review (the tools are the arbiter). This is a RASE-pattern increment: the agent realizes the hierarchy capability through the meta-harness, intrinsically verified. The lineup renders the result as per-term Broader/Narrower navigation. (2) Atlas glossary-sync — the AtlasGlossaryTerm/Category projector (extending the existing rdbms_* relational projector), a suggestion-returning projection keyed on qualifiedName, carrying the same SKOS annotations as term attributes.

Methodology — factored gates over a scaling ladder

The programme is a high-dimensional, multi-objective optimization decomposed into a sequence of low-DOF gates, valid because the factorization respects the problem’s interaction structure:

• 1-DOF where separable (architecture ⊥ data-mix; scale is a ladder, not a competing knob).
• 2-factor where the interaction is the hypothesis — specifically α (ontology-corpus fraction) × β (SQL/DDL fraction): semantics × syntax of relations, the most likely super-additive effect.
• Report Pareto slices, don’t scalarize — each gate’s output is a frontier (general vs relational vs DE-elucidation); the operating point is chosen once the surface is mapped, not at gate 1.
• No scaled spend without a green gate (the standing EVIDENCE rule). Each milestone is a pre-registered EVIDENCE entry; a failure localizes to a named arrow, not a diffuse “it didn’t work.”

Component map (the eventual GitHub project)

Milestones

Status (2026-07-09). M0 remains the last closed milestone; M1 is the gated next spend (GPU), with Path A (World-v3 augmentation: continue-pretrain vanilla RWKV-7 0.19B ± the grounded mix) locked as the data-value harness whose calibrated α Path B inherits. Everything landed since M0 sits on the instrument side, under the no-scaled-spend rule: the corpus pipeline became one idempotent flow (just metaflow → src/aegir/flows/sdg_corpora_flow.py; the pipeline chapter); Track A relational realism shipped with the token-yield bound measured early (Y_eff ≈ 7.9M — M2’s “bound to quantify”); the refinement loop closed, scaled, and staged the v0.4 corpus (refinement loop); the realized ontology publishes under the OQuaRE/IOF hard gate; and the inverted topic layer passed its phase gate — R1’s topic instrument is live and BERTopic-free (gate record).

M0 — Foundation (DONE, 2026-06-16). The substrate-evolution machinery: HermiT coherence gate (inc-2a), single-file harness (inc-2b), Meta-Harness outer loop + first discovered harness (inc-2c), realization-as-CPA beachhead (inc-2d). Committed; see EVIDENCE.md. The reasoner gates and computes; the harness evolves.

M1 — Architecture baseline (the H-Net isolation gate). Train H-Net+RWKV on RWKV-7’s open corpus, swapping only the tokenizer for byte-level dynamic chunking; establish parity across the scaling ladder to RWKV-7-matched params. DOF = 1 (the chunking change), corpus held constant. Gate: H-Net+RWKV ≥ RWKV-7 at matched scale on standard evals → architecture certified, isolated.

M2 — Instrument validity (the decisive corpus gate + proxy calibration). A same-architecture, matched-budget matrix at ≥2 ladder rungs:

• arms: grounded mix / no-ontology ablation / standard-only (the ablation arm already exists).
• α×β 2-factor cell replicated at two scales — one design yields both the interaction sign and the scale-drift of the mix optimum.
• eval: FLOOR (grounded ≈ standard on general LM evals → non-degeneracy, the failure mode every upstream proxy is blind to) + LIFT (grounded > ablation on relational + DE-elucidation; cells-only, control tasks, PR-metrics, bootstrap CI).
• side-product: calibrate the cheap proxies (R1 / coverage-close / corpus-quality) against the pretrain signal — the E1 / RASE calibration loop, run once to certify the proxies that drive iteration.
• bound to quantify: the corpus’s max non-repetitive token yield (caps α at scale). Gate: floor held AND lift CI-clean AND α×β interaction + scale-drift characterized.

M3 — Scale + the Final Phase Gate. Conditioned on M2 green: climb to RWKV-7-matched params, extrapolate α*(N)/β*(N) to target scale, confirm the lift persists. (Gate text below.)

M4 — The forward door (unlocked by the final gate). Iterate depth/breadth of the machinery: (a) close the generator loop — SAE-instrumented Qwen with process-reward fine-tuning (anti-Goodhart: reward the ontological-reasoning circuit, not just the verdict), the mutually-affirming ontology↔generator cycle anchored by the downstream model; (b) RASE in a novel domain — apply the pipeline to a second information domain with minimal re-tuning and measure what breaks (the topic layer/lexicon, the aperture, BFO anchoring, congruence/R1). Promotes “valid instrument” → “validated method.”

Final phase gate

At RWKV-7-matched scale, the ontology-grounded data mix yields an H-Net+RWKV model that (a) matches RWKV-7 on general/standard evals — the non-degeneracy floor — AND (b) exceeds the no-ontology-ablation control on relational understanding + de novo Data Element elucidation, CI-clean, with the α×β interaction and the mix-optimum scale-drift characterized.

Passing certifies the ontology machinery as a valid instrument for relational domain adaptation and authorizes the RASE-generalization phase (M4 → novel domains). Failing localizes the break to a named arrow — ontology→corpus degeneracy (M2 floor), corpus→model transfer (M2 lift), or scale-drift (M3) — each of which has its own remediation. This gate supersedes the proxy-only corpus-as-deliverable gate: the proxies are calibrated by it, not trusted ahead of it.

Three external anchors (held live throughout)

The symbolic co-evolution (ontology ↔ generator ↔ harness) optimizes formal + proxy signals only; three non-co-adapting anchors keep it honest: the reasoner (formal ground truth), the corpus (empirical fit), and the held-out H-Net+RWKV (behavioral domain truth). No milestone closes on proxies alone.