core/docs/research
Shay c98f1e07b2 feat(deduction-serve): Phase 3 — earned SERVE license via reliability gate (ADR-0256)
Deduction serving stops being a bare flag and becomes an EARNED capability
governed by the ADR-0175 reliability gate over the ADR-0199 learning arena
-- the arena's second concrete instance (GSM8K math is the first) and the
reliability substrate's first non-estimation serving consumer, a concrete
dent in that zone's standing 'designed, not wired' critique.

Non-circular thesis: the ROBDD engine is sound+complete (never wrong on the
problem it's handed), so the license does NOT certify it. It certifies the
full pipeline (reader -> projector -> engine) per argument shape -- the
FALLIBLE part is the template reader, which can misparse an argument and
hand the engine the wrong problem. The arena catches that as a 'wrong' by
comparing the pipeline's committed outcome to by-construction gold.

New: generate/proof_chain/shape.py (4 exhaustive structural shape-bands, the
capability axis, shared by arena + serving), evals/deduction_serve/practice/
(the deduction arena instance: deterministic synthetic corpus, by-construction
gold independent of the reader per ADR-0199 L-2, cross-checked against the
INDEPENDENT truth-table oracle before sealing), chat/data/deduction_serve_
ledger.json (committed SHA-sealed ledger, 4 bands x 720 correct/0 wrong),
chat/deduction_serve_license.py (tamper-evident serving reader, mirrors
estimation_license.py), tests/test_deduction_serve_license.py (13 tests),
docs/adr/ADR-0256 (+ resolves the ADR-0206 numbering collision in doc form).

Changed: chat/deduction_surface.py (composer consults the license: earned
band -> authoritative Phase-1 surface; unearned/stripped ledger -> same sound
answer served DISCLOSED/hedged -- authority now rests on committed evidence,
not a boolean), core/config.py (flag docstring: enables an EARNED path),
core/cli_test.py (deductive suite runs the license test).

All four structural bands earn SERVE at reliability 0.99087 (>= theta_SERVE
0.99) with wrong=0, so Phase-1 behavior is preserved byte-for-byte; the
gate's teeth are proven by injecting an empty ledger (the same sound answer
degrades to a disclosed hedge). Did NOT reuse the ADR-0206 govern_response
bridge (its STRICT/APPROXIMATE 'widen-past-strict' semantics are the opposite
shape from deduction's always-sound answer); did NOT rewrite report.json's
stale adr field (SHA-pinned bytes, documented in ADR-0256 s2a instead).

[Verification]: smoke 180 passed; cognition 122 passed/1 skipped;
core test --suite deductive 38 passed; architectural_invariants 75 passed;
practice runner 4 bands all SERVE wrong=0.
2026-07-23 12:59:50 -07:00
..
0148-first-conversion-proof.md docs(reader-arc): recalibration ADR-0251 + preserved knowledge + overfit inventory 2026-07-19 13:42:47 -07:00
adr-0241-0242-adversarial-and-fidelity-findings.md docs(research): W1 adversarial findings + 4-blueprint integration plan 2026-07-15 12:49:00 -07:00
adr-0241-0242-blueprint-integration-plan.md docs: close ADR-0242 memo fidelity slice + living-summary update + sanctioned-seam note (follow-up batch) 2026-07-15 15:10:31 -07:00
ADR-0242-deterministic-fibonacci-operators-and-evidence-gated-optimization.md docs(analysis): rejected Drive-draft re-implementation — assessment + sketch-defect pins 2026-07-16 10:13:04 -07:00
ADR-0243-wave-field-cognitive-lifecycle-comprehension-reasoning-and-resonant-learning.md docs(governance): weekly-audit rulings T1/T7/T8/T9 — INV-32/33/34 registration, ADR-0243 fold, residual escalation, path sweep, prune ledger 2026-07-22 19:18:59 -07:00
compare-increment-funnel-2026-07-18.md revert(reader-arc): hold frame-anchoring out of increment 1 — defer to rate-frame increment (#77 ruling) 2026-07-18 20:03:38 -07:00
compare-multiplicative-increment-plan-2026-07-18.md docs(reader-arc): fold PR #76 amendments into compare increment plan; X=15 ratified 2026-07-18 16:07:41 -07:00
CORE-ASI-Super-Blueprint-Third-Door-Horizon.md fix(third-door): align with perfected package contracts from Downloads artifacts (refs #10 #11 #12 #13) 2026-07-11 22:05:02 -07:00
core_labs_core_technical_gap_audit.pdf docs(research): intelligence-loop arc opening record — plan + Spark-audit adjudication 2026-07-18 08:14:29 -07:00
core_repository_systems_audit_and_gap_analysis.pdf docs(research): intelligence-loop arc opening record — plan + Spark-audit adjudication 2026-07-18 08:14:29 -07:00
deduction-serve-arc-phase1-turn-spine-2026-07-23.md feat(deduction-serve): Phase 1 — deduction turn spine (flag-gated, propositional Band v1) 2026-07-23 12:14:28 -07:00
deduction-serve-arc-phase2-eval-lane-2026-07-23.md feat(deduction-serve): Phase 2 — end-to-end eval lane, SHA-pinned wrong=0 gate 2026-07-23 12:36:59 -07:00
deduction-serve-arc-phase3-earned-license-2026-07-23.md feat(deduction-serve): Phase 3 — earned SERVE license via reliability gate (ADR-0256) 2026-07-23 12:59:50 -07:00
dev2-frontier-measurement-spike-2026-07-18.md docs(reader-arc): fold PR #76 amendments into compare increment plan; X=15 ratified 2026-07-18 16:07:41 -07:00
epistemic_state_math_audit_findings.md docs(epistemic): align audit terminology with runtime enum 2026-05-28 04:17:19 -07:00
field-reasoner-wedge-selection.md docs(research): select quantitative field reasoner wedge 2026-06-04 17:23:57 -07:00
increment-1-band-plan-2026-07-19.md docs(reader-arc): band-solve=0 outcome + minimal-convertible-band measurement 2026-07-18 21:37:16 -07:00
independent-comprehension-agreement-gate.md docs(research): define independent comprehension agreement gate 2026-06-04 17:23:20 -07:00
intelligence-loop-homestretch-plan-2026-07-18.md docs(research): plan record — mark EXECUTED/merged, add missed Phase-4 DONE marker 2026-07-18 11:27:02 -07:00
paradigm-deadcode-audit-2026-07-19.md docs(paradigm): dead-code audit — zero removals, all candidates traced and verified 2026-07-19 15:15:52 -07:00
reader-arc-overfit-inventory-2026-07-19.md docs(reader-arc): recalibration ADR-0251 + preserved knowledge + overfit inventory 2026-07-19 13:42:47 -07:00
reader-arc-recalibration-preserved-knowledge-2026-07-19.md docs(reader-arc): recalibration ADR-0251 + preserved knowledge + overfit inventory 2026-07-19 13:42:47 -07:00
reader-hamiltonian-compiler-spike-2026-07-18.md feat(adr-0249): P1 conformal quantity kernel 2026-07-18 12:04:28 -07:00
robdd_proof_chain_design_note.md Add proof-chain corpus fixtures 2026-06-02 19:26:46 -07:00
spark-audit-adjudication-2026-07-18.md docs(research): intelligence-loop arc opening record — plan + Spark-audit adjudication 2026-07-18 08:14:29 -07:00
third-door-blueprint-fidelity.md docs(governance): RATIFY — ADR-0241 and ADR-0242 → Accepted (Joshua Shay, 2026-07-15) 2026-07-15 15:25:47 -07:00
THIRD-DOOR-LANDING-PACKAGE-README.md fix(third-door): align with perfected package contracts from Downloads artifacts (refs #10 #11 #12 #13) 2026-07-11 22:05:02 -07:00
tier2-multi-entity-arithmetic-spike-2026-07-18.md docs(research): Tier-2 multi-entity arithmetic design spike 2026-07-18 14:01:15 -07:00
trackb-inc2-generalization-2026-07-20.md fix(trackb): pure-S1 extract refuse multi-clause overmatch (0361 wrong=0) 2026-07-19 20:04:27 -07:00
trackb-s1-slice-2026-07-20.md fix(trackb): pure-S1 gate — refuse total supersets and extra contains 2026-07-19 19:10:05 -07:00