← Archive
lm-003856 · 2026-09

自主數學研究 Runtime:從可定址認知到自主理論建構

下載 MD 檔 ⬇

自主數學研究 Runtime:從可定址認知到自主理論建構

Autonomous Mathematical Research Runtime: From Addressable Cognition to Self-Directed Theory Construction

系列:Autonomous Mathematical Research / Paper 04 of 04
版本:v0.1
日期:2026-08-23
作者:Neo.K


摘要

前三篇依序建立了自主數學研究的統合理論、多域問題診斷與受約束數學域補全,以及問題身份、理論擴張與數學義務框架。然而,若這些概念只停留在論文語義層,它們仍不足以形成真正可運行的自主數學研究 AI。本文提出 Autonomous Mathematical Research Runtime, AMRR(自主數學研究 Runtime) 的第一版可執行架構,目標是把數學研究中的問題、理論、缺口、修復、義務、證明、反例、方法、橋接與新穎性狀態轉換為 canonical runtime objects,並將它們接入既有 Addressable Cognitive Runtime(ACR)與 CTCL-ITR 時間因果證據層。

本文不重寫 ACR。現有 ACR 已具有 canonical schema、可定址 CognitiveObject、Semantic State Encoder、Cognitive Affordance、Cognitive Program、Governance、Decision Receipt 與 Persistent Loop 的架構邊界;AMRR 在此基礎上加入數學 domain extension。本文提出九個主要數學模組:Mathematical State Adapter、Problem Store、Theory Store、Gap Diagnosis Engine、Constrained Repair Engine、Mathematical Obligation Engine、Mathematical Cognitive Registry、Verifier Router / Verifier Ensemble、Theory Bridge & Novelty Layer,並以 Research Runtime Coordinator 將它們編排為持續研究閉環。

AMRR 的 canonical state 不再只保存「目前答案」,而保存 ProblemVersionProblemVersionTheoryVersionTheoryVersionGapMapGapMapRepairCandidateRepairCandidateMathChangeSetMathChangeSetAssumptionEnvelopeAssumptionEnvelopeObligationGraphObligationGraphVerificationReceiptVerificationReceiptTheoryBridgeTheoryBridgeResearchReceiptResearchReceipt。任何 AI 生成的定義、假設、lemma、方法、猜想或公理都先進入 candidate state;只有經過 Problem Identity、Theory Extension、Obligation、Verification 與 Governance gates 後,才能 promotion 到 accepted research state。本文同時定義 cog://math/* 認知算子 namespace、typed verifier selection、problem/theory DAG、數學 receipts、CTCL event namespace、研究契約、budget model、failure model、multi-agent join 與 context compression 邊界。

本文最後提出一條由現有 ACR Phase 2 向 AMRR MVP 演進的實作路線:保留 Phase 0–2 的 frozen interfaces,先新增 Math Extension Schemas 與 deterministic Mathematical State Adapter,再依序建立 Gap Diagnosis、Problem Identity / Theory Store、Obligation Engine、Math Cognitive Registry、Verifier Router、CMDC Repair Engine、Theory Bridge、CTCL Adapter 與 Persistent Research Loop。每一層皆遵守:

ArchitectureCapabilityFalsificationGate\boxed{ Architecture \rightarrow Capability \rightarrow FalsificationGate }

而不是以「AI 看起來很會研究」作為完成標準。

本文的核心命題是:自主數學研究 AI 不應被實作為一個無界循環的文字 Agent,而應被實作為一個以 canonical mathematical state、typed cognitive programs、受約束 theory mutation、外部 verifier 與可驗證時間因果歷史共同構成的 persistent research runtime。

關鍵詞: Autonomous Mathematical Research Runtime、AMRR、Addressable Cognitive Runtime、CMDC、數學狀態、問題身份、理論擴張、數學義務、形式驗證、Theory Bridge、CTCL-ITR


1. 問題:前三篇已經有方法論,但還沒有 Runtime

前三篇可以壓縮成:

ResearchStateDiagnoseRepairIdentity/ExtensionObligationVerify.ResearchState \rightarrow Diagnose \rightarrow Repair \rightarrow Identity/Extension \rightarrow Obligation \rightarrow Verify.

這仍然是一條概念鏈。

真正工程問題是:

哪些東西是 canonical object?誰保存它?誰可以改?誰驗證?哪些狀態可以 promotion?錯誤如何回退?上下文被壓縮後如何恢復?

因此本文研究:

MethodologyExecutableRuntimeArchitecture.\boxed{ Methodology \rightarrow ExecutableRuntimeArchitecture. }

2. AMRR 的設計原則

AMRR v0.1 固定十個工程原則。

第一:

Integrate>Rewrite.\boxed{ Integrate > Rewrite. }

第二:

CanonicalStateRenderedText.\boxed{ CanonicalState \neq RenderedText. }

第三:

CandidateAcceptedState.\boxed{ Candidate \neq AcceptedState. }

第四:

ProblemTheoryProofClaim.\boxed{ Problem \neq Theory \neq Proof \neq Claim. }

第五:

DiagnosisRepairVerification.\boxed{ Diagnosis \neq Repair \neq Verification. }

第六:

CanGenerateMayAdoptMayClaim.\boxed{ CanGenerate \neq MayAdopt \neq MayClaim. }

第七:

WorkingContextResearchLedger.\boxed{ WorkingContext \neq ResearchLedger. }

第八:

ModelRoleModelIdentity.\boxed{ ModelRole \neq ModelIdentity. }

第九:

VerificationSingleVerifier.\boxed{ Verification \neq SingleVerifier. }

第十:

ClaimScopeVerifiedScope.\boxed{ ClaimScope \not> VerifiedScope. }

3. Implemented Baseline 與 Proposed Extension 必須分開

AMRR v0.1 不把尚未實作的模組描述成既有功能。

目前 ACR 工程基線已具有:

Phase 0: canonical schema freeze
Phase 1: SPRC / CIO adapter and addressable CognitiveObjects
Phase 2: deterministic public Semantic State Encoder

Phase 2 artifact 的 release validation 記錄包含 10 個 Phase-0 schemas、64 個 Phase-1 CognitiveObjects、deterministic SemanticState fingerprint 與完整 release suite 的 73 passed

本文新增的:

Mathematical State Adapter
Problem Store
Theory Store
Gap Diagnosis Engine
Repair Engine
Obligation Engine
Math Cognitive Registry
Verifier Router
Theory Bridge Layer
Persistent Research Loop

均屬 AMRR Proposed Extension


4. 總體架構

AMRR 可以表示為:

AMRR=ACR+MathDomain+CMDC+VerifierLayer+TheoryBridge+CTCL-ITR.\boxed{ AMRR = ACR + MathDomain + CMDC + VerifierLayer + TheoryBridge + CTCL\text{-}ITR. }

邏輯上分為五層:

Layer A — Mathematical State & Knowledge
Layer B — Cognitive Research Runtime
Layer C — Mutation / Obligation / Verification
Layer D — Governance & Research Contract
Layer E — Temporal-Causal Evidence & Audit

5. Layer A — Mathematical State & Knowledge

Layer A 回答:

現在的數學世界狀態是什麼?

核心物件:

MathematicalState
ProblemRecord
ProblemVersion
TheoryRecord
TheoryVersion
DefinitionRecord
AssumptionRecord
ClaimRecord
MethodRecord
GapMap
ObligationGraph
BridgeGraph

這一層不負責自由生成研究方案。

它負責:

Store,Normalize,Version,Resolve,Reference.Store, Normalize, Version, Resolve, Reference.

6. Layer B — Cognitive Research Runtime

Layer B 回答:

現在應該怎麼研究?

核心:

Mathematical Affordance Retriever
Math Cognitive Router
Research Program Compiler
Research Program Executor
Re-observer
Agenda Runtime

它使用:

StMS_t^M

取得:

AM(StM).\mathcal A_M(S_t^M).

再編譯:

PtM.P_t^M.

7. Layer C — Mutation / Obligation / Verification

Layer C 回答:

這個新東西能不能合法進入 theory state?

核心:

Gap Diagnosis Engine
Constrained Repair Engine
Problem Identity Resolver
Theory Extension Classifier
Mathematical Obligation Engine
Verifier Router
Verifier Adapters
Promotion Engine

其主要輸出不是 prose。

而是:

MathChangeSet,ObligationSet,VerificationReceipt,PromotionDecision.MathChangeSet, ObligationSet, VerificationReceipt, PromotionDecision.

8. Layer D — Governance & Research Contract

Layer D 回答:

AI 是否被允許自行採用、公開或升格這項數學變更?

沿用:

Decision{EXECUTE,REFUSE,DEFER,IDLE,ESCALATE}.Decision \in \{ EXECUTE, REFUSE, DEFER, IDLE, ESCALATE \}.

數學域映射成:

CanGenerate(x),CanGenerate(x), ShouldPursue(x),ShouldPursue(x), MayAdopt(x),MayAdopt(x), MayClaim(x).MayClaim(x).

9. Layer E — Temporal-Causal Evidence & Audit

Layer E 回答:

這個問題何時被改了?為什麼改?當時知道什麼?哪些 proof / verifier / assumption 支撐現在的 claim?

核心:

CTCL instant
TemporalEvent
DecisionReceipt
ProblemMutationReceipt
TheoryExtensionReceipt
DiagnosisReceipt
RepairReceipt
ObligationReceipt
VerificationReceipt
GapClosureReceipt
ResearchReceipt
CommitReceipt

因此:

ResearchHistoryConversationHistory.\boxed{ ResearchHistory \neq ConversationHistory. }

10. Mathematical State 不應取代原 Semantic State

ACR Phase 2 已經有:

SemanticStatet.SemanticState_t.

AMRR 不應建立另一套互不相容的根狀態。

而應:

MathematicalStatet=Extend(SemanticStatet,MathExtensiont).\boxed{ MathematicalState_t = Extend( SemanticState_t, MathExtension_t ). }

因此通用欄位繼續使用:

progress
uncertainty
failures
missing
risks
budget
authority_ref
goal_refs
commitments
environment_refs
memory_refs

數學欄位進:

extensions.math

11. extensions.math v0.1

第一版可以定義:

{
  "problem_refs": [],
  "active_problem_ref": null,
  "theory_refs": [],
  "active_theory_ref": null,

  "definition_refs": [],
  "assumption_refs": [],
  "method_refs": [],
  "claim_refs": [],

  "gap_map_ref": null,
  "open_obligation_refs": [],
  "verification_refs": [],
  "bridge_refs": [],

  "novelty_state": "unknown",
  "research_phase": "diagnosis"
}

這個 extension 必須 canonical serialize。


12. Mathematical State Identity

狀態 identity 不用自然語言摘要。

沿用 canonical fingerprint:

StateID=Hash(CanonicalSerialize(StM)).StateID = Hash( CanonicalSerialize(S_t^M) ).

因此:

EquivalentNormalizedStateSameFingerprint.EquivalentNormalizedState \rightarrow SameFingerprint.

這保持 Phase 2 deterministic boundary。


13. Problem Store

Problem Store 保存:

ProblemRecord.ProblemRecord.

最小 schema:

{
  "problem_id": "problem:...",
  "version": 3,
  "parent_refs": [],
  "canonical_statement_ref": "artifact:...",
  "informal_statement_ref": "artifact:...",
  "formal_statement_refs": [],
  "domain_refs": [],
  "success_criteria": [],
  "status": "active",
  "created_from_gap_ref": null,
  "relation_to_parent": null,
  "assumption_envelope_ref": "assumption-envelope:..."
}

14. Problem Version Graph

問題不能只靠:

problem:v7

還需要:

GP=(VP,EP).G_P = (V_P,E_P).

edge:

PirPj,P_i \xrightarrow{r} P_j,

其中:

rRP.r \in \mathcal R_P.

這使:

Restriction,Generalization,AddedAssumption,NewProblemRestriction, Generalization, AddedAssumption, NewProblem

都可以共存於 Problem DAG。


15. Problem Relation Resolver

接口:

problem.relate(before_ref, after_ref)

輸出:

{
  "relation": "restriction",
  "confidence": 0.93,
  "evidence_refs": [],
  "verification_status": "candidate"
}

relation 也不是 oracle。

所以:

CandidateRelationVerifiedRelation.CandidateRelation \neq VerifiedRelation.

16. Theory Store

Theory Store 保存:

TheoryRecord.TheoryRecord.

最小:

{
  "theory_id": "theory:...",
  "version": 4,
  "parent_refs": [],
  "signature_refs": [],
  "axiom_refs": [],
  "definition_refs": [],
  "theorem_refs": [],
  "method_refs": [],
  "bridge_refs": [],
  "open_obligation_refs": [],
  "extension_profile_ref": null
}

17. Theory State 不等於 Document

保持:

TheoryStatePaperText.\boxed{ TheoryState \neq PaperText. }

論文、README、自然語言摘要只是 renderer / artifact。

Canonical theory state 應可被機器操作。


18. Theory Extension Classifier

接口:

theory.classify_extension(before, after)

輸出:

ExtensionProfile.ExtensionProfile.

例如:

{
  "language_change": true,
  "semantic_change": false,
  "proof_strength_change": false,
  "model_change": false,
  "foundation_change": false,
  "class": "definitional_extension"
}

19. Gap Diagnosis Engine

輸入:

StM.S_t^M.

輸出:

Δt.\Delta_t.

第一版不要直接全交給 LLM。

可以採:

RuleEvidence+ToolEvidence+ModelClassification+CrossCheck.\boxed{ RuleEvidence + ToolEvidence + ModelClassification + CrossCheck. }

20. Gap Diagnosis 的資料源

允許:

formalizer errors
proof assistant errors
counterexamples
failed method history
missing library symbol
ambiguous definitions
problem relation evidence
verification receipts
literature / prior-art refs
compute budget exhaustion
human review annotations

禁止把 private chain-of-thought 當 canonical evidence。


21. Gap Map Schema

{
  "gap_map_id": "gap-map:...",
  "state_ref": "state:...",
  "gaps": [
    {
      "gap_id": "gap:...",
      "domain": "definition",
      "target_ref": "definition:...",
      "evidence_refs": [],
      "confidence": 0.91,
      "severity": "high",
      "blocking": true,
      "depends_on": [],
      "candidate_repair_classes": []
    }
  ]
}

22. Gap Map 是 DAG,不只是 list

因:

DefinitionGapFormalizationGapProofGap.DefinitionGap \rightarrow FormalizationGap \rightarrow ProofGap.

因此:

GΔ=(VΔ,EΔ).G_\Delta = (V_\Delta,E_\Delta).

這可以直接利用 CTCL / topology style 的 multi-parent graph semantics。


23. Gap Prioritizer

Priority(δi)=f(Blocking,Severity,Confidence,ExpectedGain,RepairCost,DependencyCentrality).Priority(\delta_i) = f( Blocking, Severity, Confidence, ExpectedGain, RepairCost, DependencyCentrality ).

第一版不應將其壓成不可解釋 reward。

可以保留 vector:

Score(δi)=(B,S,C,G,K,D).Score(\delta_i) = ( B,S,C,G,K,D ).

24. Constrained Repair Engine

輸入:

δi+StM+ResearchContract.\delta_i + S_t^M + ResearchContract.

輸出:

R(δi).\mathcal R(\delta_i).

每個 repair:

rj=(Type,Target,Mutation,ExpectedEffect,Risks,Obligations).r_j = ( Type, Target, Mutation, ExpectedEffect, Risks, Obligations ).

25. Repair 不能直接 Commit

保持三層:

RepairCandidateRepairDecisionAcceptedMutation.RepairCandidate \rightarrow RepairDecision \rightarrow AcceptedMutation.

因此:

RepairCandidateTheoryMutation.\boxed{ RepairCandidate \neq TheoryMutation. }

26. MathChangeSet

{
  "change_set_id": "math-change:...",
  "source_state_ref": "state:...",
  "gap_ref": "gap:...",
  "change_type": "add_assumption",
  "target_ref": "problem:...",
  "before_ref": "problem:v3",
  "after_candidate_ref": "problem:v4-candidate",
  "reason_codes": [],
  "obligation_refs": [],
  "verification_status": "pending",
  "promotion_status": "candidate"
}

27. Mathematical Obligation Engine

接口:

obligation.generate(change_set)

輸出:

O(change).\mathcal O(change).

第一版採 rule-first:

add_definition
→ well_formedness
→ semantic_faithfulness
→ bridge_check
→ conservativity_if_claimed

add_assumption
→ problem_identity
→ scope_disclosure
→ assumption_envelope_update

new_axiom
→ explicit_nonconservative_status
→ consistency/model obligations
→ escalation policy

28. Obligation Graph

義務之間也有 dependency:

O1O2.O_1 \rightarrow O_2.

例如:

SemanticFaithfulnessClaimScopeValidation.SemanticFaithfulness \rightarrow ClaimScopeValidation.

因此:

GO=(VO,EO).G_O = (V_O,E_O).

29. Obligation Schema

{
  "obligation_id": "obligation:...",
  "class": "semantic_faithfulness",
  "trigger_ref": "math-change:...",
  "claim_ref": null,
  "required_evidence": [],
  "verifier_classes": ["formal", "semantic", "human"],
  "depends_on": [],
  "status": "OPEN",
  "waiver_policy": "escalation_required"
}

30. Obligation Status Machine

OPENPARTIALDISCHARGED.OPEN \rightarrow PARTIAL \rightarrow DISCHARGED.

也可以:

OPENFAILED.OPEN \rightarrow FAILED.

或:

OPENWAIVED.OPEN \rightarrow WAIVED.

但:

WAIVEDDISCHARGED.\boxed{ WAIVED \neq DISCHARGED. }

31. Mathematical Cognitive Registry

AMRR 不應每輪重新用自然語言發明「現在該怎麼研究」。

建立:

cog://math/interpret@1
cog://math/formalize@1
cog://math/diagnose-gap@1
cog://math/clarify-problem@1
cog://math/define@1
cog://math/add-assumption-candidate@1
cog://math/change-representation@1
cog://math/generate-lemma@1
cog://math/generate-conjecture@1
cog://math/search-counterexample@1
cog://math/retrieve-method@1
cog://math/generate-method@1
cog://math/prove@1
cog://math/disprove@1
cog://math/check-conservativity@1
cog://math/verify-faithfulness@1
cog://math/bridge-theory@1
cog://math/check-prior-art@1
cog://math/discharge-obligation@1
cog://math/package-theory@1

32. Math Cognitive Object

每個 math operator 沿用 CognitiveObject:

Identity(C)=(namespace,id,version,schemaHash).Identity(C) = (namespace,id,version,schemaHash).

再加入 domain-specific metadata:

{
  "math_domains": ["definition", "verification"],
  "accepted_gap_classes": [],
  "produces_candidate_types": [],
  "obligation_effects": [],
  "required_verifier_classes": []
}

33. Math Affordance Retrieval

輸入:

StM,Δt,Gt,Ct.S_t^M, \Delta_t, G_t, C_t.

輸出:

AM.\mathcal A_M.

評分:

Score(Ωi)=wgGapFit+wpPreconditionFit+whHistoricalUtility+woObligationReductionwcCostwrRisk.Score(\Omega_i) = w_gGapFit + w_pPreconditionFit + w_hHistoricalUtility + w_oObligationReduction - w_cCost - w_rRisk.

34. Research Program Compiler

CompilerM(StM,Δt,Gt,Ct,AM)PtM.Compiler_M( S_t^M, \Delta_t, G_t, C_t, \mathcal A_M ) \rightarrow P_t^M.

例如:

{
  "program_id": "math-program:...",
  "steps": [
    {"op": "cog://math/search-counterexample@1"},
    {"op": "cog://math/clarify-problem@1"},
    {"op": "cog://math/formalize@1"},
    {"op": "cog://math/prove@1"}
  ],
  "stop_if": [
    "counterexample_found",
    "problem_identity_ambiguous",
    "budget_exhausted"
  ]
}

35. Program Validation

執行前:

TypeCheckScopeCheckAuthorityCheckBudgetCheckProblemIdentityCheckInvariantCheck.TypeCheck \land ScopeCheck \land AuthorityCheck \land BudgetCheck \land ProblemIdentityCheck \land InvariantCheck.

若失敗:

ProgramRejected.ProgramRejected.

36. Verifier Router

不是每個 obligation 都丟給 Lean。

定義:

VerifierRouter(Claim,Obligation,TheoryState){V1,,Vk}.VerifierRouter( Claim, Obligation, TheoryState ) \rightarrow \{V_1,\ldots,V_k\}.

37. Verifier Ensemble

第一版 class:

formal_proof
symbolic_algebra
sat
smt
numeric
exhaustive_search
simulation
semantic_equivalence
literature_prior_art
human_review

因此:

Verification=TypedVerifierSelection.\boxed{ Verification = TypedVerifierSelection. }

38. Verifier Adapter Interface

統一接口:

verify(request) -> VerificationResult

request 至少帶:

claim_ref
theory_ref
problem_ref
obligation_ref
assumption_envelope_ref
budget

結果:

PASS
FAIL
INCONCLUSIVE
ERROR

而:

INCONCLUSIVEFAIL.INCONCLUSIVE \neq FAIL.

39. Verification Receipt

{
  "verification_receipt_id": "verify:...",
  "claim_ref": "claim:...",
  "obligation_ref": "obligation:...",
  "verifier": {
    "class": "formal_proof",
    "implementation": "lean"
  },
  "theory_ref": "theory:v4",
  "assumption_envelope_ref": "ae:...",
  "result": "PASS",
  "artifact_refs": [],
  "executed_at": "...",
  "knowledge_boundary_ref": "kb:..."
}

40. Proof Result 不能自動成為 Claim

保持:

ProofFoundFormalVerifiedSemanticallyFaithfulClaimApproved.ProofFound \neq FormalVerified \neq SemanticallyFaithful \neq ClaimApproved.

因此 promotion pipeline:

CandidateClaimVerificationScopeCheckObligationCheckGovernanceAcceptedClaim.CandidateClaim \rightarrow Verification \rightarrow ScopeCheck \rightarrow ObligationCheck \rightarrow Governance \rightarrow AcceptedClaim.

41. Assumption Envelope Service

接口:

claim.assumption_envelope(claim_ref)

計算:

AE(C)=DependencyClosure(C)Assumptions.AE(C) = DependencyClosure(C) \cap Assumptions.

它必須跨:

claim
lemma
definition
method
theory bridge
formalization

追蹤依賴。


42. No-Laundering Gate

公開 claim 前:

ClaimScope(C)VerifiedScope(C)ClaimScope(C) \subseteq VerifiedScope(C)

以及:

ClaimAssumptions(C)AE(C).ClaimAssumptions(C) \supseteq AE(C).

若不成立:

ClaimBlocked.ClaimBlocked.

43. Theory Bridge Layer

Theory Bridge 回答:

新定義、新理論與既有數學到底如何連接?

Bridge object:

B=(Source,Target,Relation,Evidence,Strength).B = (Source,Target,Relation,Evidence,Strength).

relation:

equivalent_to
generalizes
specializes
reduces_to
interprets
isomorphic_to
independent_from
uses
contradicts

44. Bridge Strength

Strength(B){Exact,Formal,Partial,SemanticCandidate,Heuristic}.Strength(B) \in \{ Exact, Formal, Partial, SemanticCandidate, Heuristic \}.

下游使用 bridge 時必須知道 strength。

因此:

Heuristic⇏FormalEquivalence.Heuristic \not\Rightarrow FormalEquivalence.

45. Prior-Art / Novelty Layer

Novelty 不應是 boolean。

第一版:

unknown
search_incomplete
known_prior_art
rediscovered
novel_candidate
human_review_required

因此:

NovelCandidate(XKt).NovelCandidate(X\mid K_t).

而不是:

AbsolutelyNovel(X).AbsolutelyNovel(X).

46. Discovery Receipt

{
  "discovery_receipt_id": "discovery:...",
  "contribution_ref": "claim:...",
  "knowledge_boundary_ref": "kb:...",
  "search_scope": [],
  "prior_art_refs": [],
  "novelty_status": "novel_candidate",
  "review_status": "pending"
}

47. Research Contract

AMRR 的持續輸入不是一個 prompt。

而是:

ResearchGoal+ResearchEnvironment+ResearchContract.\boxed{ ResearchGoal + ResearchEnvironment + ResearchContract. }

第一版 contract:

CR=(Scope,MutationPolicy,VerificationPolicy,ComputeBudget,LiteraturePolicy,ClaimPolicy,EscalationPolicy,Termination).C_R = ( Scope, MutationPolicy, VerificationPolicy, ComputeBudget, LiteraturePolicy, ClaimPolicy, EscalationPolicy, Termination ).

48. Research Contract 範例

Goal:
Investigate the target conjecture and produce a reviewable theory package.

Allowed:
- generate definitions, lemmas, conjectures and methods as candidates
- run formal / symbolic / numerical verifiers
- change representations
- search prior art
- create problem branches

Approval required:
- replace the canonical problem
- promote a new axiom
- declare novelty
- mark package submission-ready

Denied:
- erase prior problem versions
- report restricted results as unrestricted
- hide assumptions
- fabricate citations

49. Research Agenda Runtime

輸入:

Goal+StM+GapMap+OpenObligations+Budget.Goal + S_t^M + GapMap + OpenObligations + Budget.

輸出:

AgendaCandidate.AgendaCandidate.

例如:

investigate_definition_gap
search_counterexample
formalize_candidate_lemma
discharge_faithfulness_obligation
check_prior_art
reopen_closed_gap

50. Agenda 不能直接執行

保持:

AgendaCandidateGovernanceAgendaAccepted.AgendaCandidate \rightarrow Governance \rightarrow AgendaAccepted.

因此:

ResearchIdeaResearchAuthorization.\boxed{ ResearchIdea \neq ResearchAuthorization. }

51. Persistent Research Loop

核心偽代碼:

while runtime_active:

    observation = observe_research_environment()
    state = encode_mathematical_state(observation)

    gaps = diagnose(state)
    obligations = refresh_open_obligations(state)

    agenda_candidates = propose_research_agenda(
        goal,
        state,
        gaps,
        obligations,
        contract
    )

    if no_positive_candidate(agenda_candidates):
        enter_idle_or_stop()
        continue

    for agenda in agenda_candidates:

        agenda_decision = govern_agenda(agenda, contract)

        if agenda_decision != EXECUTE:
            record_nonexecute_decision()
            continue

        affordances = retrieve_math_affordances(state, gaps)
        program = compile_math_program(state, agenda, affordances)

        validate_program(program)
        result = execute_math_program(program)

        if result.proposes_mutation:
            change = create_math_change_set(result)
            relation = classify_problem_relation(change)
            extension = classify_theory_extension(change)
            new_obligations = generate_obligations(change)

            mutation_decision = govern_mutation(
                change,
                relation,
                extension,
                new_obligations,
                contract
            )

            if mutation_decision == EXECUTE:
                candidate_state = apply_candidate(change)
                verification = route_and_run_verifiers(candidate_state)
                promotion = evaluate_promotion(candidate_state, verification)
                commit_or_reject(promotion)

        write_receipts()
        emit_ctcl_events()
        reobserve()
        rediagnose()
        compact_context_if_needed()

52. Re-observation 是研究核心

一次 repair 後不能假設問題就好了。

必須:

StMRepairSt+1MReDiagnose.S_t^M \rightarrow Repair \rightarrow S_{t+1}^M \rightarrow ReDiagnose.

因此:

RepairClosure.\boxed{ Repair \neq Closure. }

53. Gap Reopen

已關閉 gap 可以:

ClosedReopened.Closed \rightarrow Reopened.

觸發條件:

new counterexample
new proof failure
new prior art
new semantic mismatch
new theory contradiction
new human review

這是長期研究必要能力。


54. Multi-Agent / Parallel Research

允許:

FormalProofCounterexampleSearchPriorArtSearchAlternativeRepresentation.FormalProof \parallel CounterexampleSearch \parallel PriorArtSearch \parallel AlternativeRepresentation.

最後:

Join.Join.

但 canonical state 仍由 shared stores / receipts 決定。

不能讓每個 agent 各自擁有互相矛盾的「正式問題」。


55. Multi-Agent Conflict

若:

DiagnosisADiagnosisB,Diagnosis_A \neq Diagnosis_B,

不能直接投票了事。

應生成:

DiagnosisConflict.DiagnosisConflict.

並比較:

Evidence,Calibration,VerifierResults,KnowledgeBoundary.Evidence, Calibration, VerifierResults, KnowledgeBoundary.

56. Model Roles

AMRR 邏輯角色:

Researcher
Diagnostician
Formalizer
Conjecturer
CounterexampleSearcher
MethodDesigner
VerifierController
BridgeAnalyst
Governor
Auditor

同一 model 可以扮演多角色。

保持:

RoleModelIdentity.Role \neq ModelIdentity.

57. Model Provider 不寫死

接口:

generate(request) -> response

可以是:

remote frontier model
local model
specialized theorem model
symbolic subsystem
hybrid

AMRR semantics 不綁定單一模型。


58. Tool Space 與 Cognitive Space 統一

ActionSpace=ExternalToolsCognitiveActionsMathematicalActions.ActionSpace = ExternalTools \cup CognitiveActions \cup MathematicalActions.

例如:

CALL web.search
CALL lean.verify
CALL cas.simplify
CALL cog://math/diagnose-gap@1
CALL cog://math/change-representation@1

59. Mathematical Action 也有 Authority

例如:

create conjecture candidate = ALLOW
create problem branch = ALLOW
promote new axiom = APPROVAL_REQUIRED
publish novelty claim = APPROVAL_REQUIRED
erase original problem = DENY

所以:

MathematicalCognition⇏TheoryCommit.MathematicalCognition \not\Rightarrow TheoryCommit.

60. Budget Model

AMRR budget:

B=(Tokens,Calls,WallTime,MachineTime,Money,Energy,ProofSearch,Simulation,LiteratureSearch,HumanReview).B = ( Tokens, Calls, WallTime, MachineTime, Money, Energy, ProofSearch, Simulation, LiteratureSearch, HumanReview ).

不同 operator 使用不同 budget class。


61. Budget-Aware Verification

若完整 formal proof 成本過高,可先:

CheapFalsificationExpensiveVerification.CheapFalsification \rightarrow ExpensiveVerification.

例如:

CounterexampleSearchSMTLean.CounterexampleSearch \rightarrow SMT \rightarrow Lean.

但便宜 verifier 不得被當成更強 verifier 的替代證據。


62. Failure Model

AMRR 至少區分:

problem_error
definition_error
assumption_error
representation_error
method_error
proof_error
formalization_error
verifier_error
tool_error
bridge_error
novelty_error
authority_error
protocol_error
ledger_error
budget_error
environment_error

不能全部叫:

failure

63. Unknown 是合法狀態

允許:

unknown
ambiguous
inconclusive
unresolved

因此:

UnknownFalse.Unknown \neq False.

以及:

InconclusiveRefuted.Inconclusive \neq Refuted.

64. Context Compression

AMRR working context 必然要壓縮。

但:

Compression may forget prose, but must not erase mathematical lineage.\boxed{ Compression \text{ may forget prose, but must not erase mathematical lineage.} }

必須保留:

problem refs
theory refs
gap refs
obligation refs
assumption envelopes
verification receipts
bridge refs
research receipts
causal parents

65. Research Ledger

WorkingMemoryResearchLedger.WorkingMemory \neq ResearchLedger.

Research ledger 保存:

ProblemMutation,TheoryExtension,Diagnosis,Repair,Verification,Promotion,Claim.ProblemMutation, TheoryExtension, Diagnosis, Repair, Verification, Promotion, Claim.

66. CTCL Event Namespace

第一版:

math.problem.registered
math.problem.derived
math.problem.relation.proposed
math.problem.relation.verified

math.theory.registered
math.theory.extension.proposed
math.theory.extension.classified
math.theory.extension.promoted

math.gap.detected
math.gap.updated
math.gap.closed
math.gap.reopened

math.repair.proposed
math.repair.accepted
math.repair.rejected

math.obligation.created
math.obligation.updated
math.obligation.discharged
math.obligation.waived

math.verification.requested
math.verification.completed

math.bridge.proposed
math.bridge.verified

math.claim.proposed
math.claim.approved
math.claim.blocked

math.research.idled
math.research.deferred
math.research.escalated

67. Temporal Evidence

每個重大事件至少引用:

ctcl_instant_id
occurred_at
recorded_at
run_id
interaction_round
ledger_seq
causal_parent_ids
state_ref
problem_ref
theory_ref
contract_ref
knowledge_boundary_ref

68. Research Receipt

對一個完成的研究 transition:

{
  "research_receipt_id": "research-receipt:...",
  "agenda_ref": "agenda:...",
  "state_before_ref": "state:...",
  "state_after_ref": "state:...",
  "problem_refs": [],
  "theory_refs": [],
  "gap_refs": [],
  "change_set_refs": [],
  "obligation_refs": [],
  "verification_refs": [],
  "decision_refs": [],
  "causal_parent_ids": []
}

69. Research Receipt 不保存 Private CoT

保存:

public state
selected cognitive program
reason codes
math changes
verification results
receipts
artifact refs

不要求:

hidden reasoning tokens

因此:

AuditabilityPrivateCoTLogging.Auditability \neq PrivateCoTLogging.

70. Replay 與 Re-enactment

Replay:

LedgerHistoricalResearchState.Ledger \rightarrow HistoricalResearchState.

Re-enactment:

HistoricalProgram+CurrentModelNewOutcome.HistoricalProgram + CurrentModel \rightarrow NewOutcome.

所以:

ReplayReenactment.\boxed{ Replay \neq Reenactment. }

71. AMRR API v0.1

runtime.observe_research(...)
runtime.encode_math_state(...)
runtime.diagnose_gaps(...)
runtime.propose_research_agenda(...)
runtime.retrieve_math_affordances(...)
runtime.compile_math_program(...)
runtime.execute_math_program(...)

problem.register(...)
problem.derive(...)
problem.relate(...)
problem.branch(...)

theory.register(...)
theory.extend(...)
theory.classify_extension(...)
theory.bridge(...)

repair.propose(...)
repair.apply_candidate(...)
repair.rollback(...)

obligation.generate(...)
obligation.discharge(...)
obligation.defer(...)
obligation.waive(...)

verify.route(...)
verify.run(...)

claim.assumption_envelope(...)
claim.validate_scope(...)
claim.propose(...)

research.audit(...)
research.replay(...)

72. Repo Structure 建議

不建議一開始另開完全獨立 runtime。

優先:

addressable-cognitive-runtime/
├── src/addressable_cognitive_runtime/
│   ├── state/
│   ├── registry/
│   ├── retrieval/
│   ├── compiler/
│   ├── runtime/
│   ├── governance/
│   ├── ctcl_adapter/
│   └── math/
│       ├── state/
│       ├── problem/
│       ├── theory/
│       ├── gap/
│       ├── repair/
│       ├── obligation/
│       ├── verifier/
│       ├── bridge/
│       ├── claim/
│       └── research_loop/
├── schemas/
│   └── math/
├── registry/
│   └── math/
├── adapters/
│   └── math/
├── experiments/
│   └── amrr_gates/
└── tests/
    └── math/

73. Math Schema Catalog

建議新增:

mathematical-state-extension.schema.json
problem-record.schema.json
problem-relation.schema.json
problem-mutation.schema.json
assumption-envelope.schema.json

theory-record.schema.json
theory-extension.schema.json
math-change-set.schema.json

gap.schema.json
gap-map.schema.json
repair-candidate.schema.json

mathematical-obligation.schema.json
obligation-graph.schema.json

verification-request.schema.json
verification-receipt.schema.json

theory-bridge.schema.json
discovery-receipt.schema.json
research-receipt.schema.json
claim-record.schema.json

74. Schema Versioning

每個 schema:

schema_id
schema_version
canonicalization_version

Identity:

(namespace,id,version,schemaHash).(namespace,id,version,schemaHash).

不能靠檔名推斷語義。


75. Phase M0 — Math Schema Freeze

先建立全部 math schemas。

成功條件:

round-trip
invalid-case rejection
stable IDs
canonical serialization
hash stability
no hidden implicit fields

76. Phase M1 — Mathematical State Adapter

輸入仍是公開 observation。

輸出:

SemanticState+extensions.math.SemanticState + extensions.math.

成功條件:

EquivalentObservationByteIdenticalMathState.EquivalentObservation \rightarrow ByteIdenticalMathState.

77. Phase M2 — Problem / Theory Stores

建立:

Problem DAG
Theory version graph
Assumption Envelope
stable refs

成功條件:

NoSilentOverwrite.NoSilentOverwrite.

78. Phase M3 — Gap Diagnosis Engine

先做 controlled benchmark。

第一批只支援:

definition
assumption
representation
lemma/dependency
compute
verification

不要一開始硬做全部十三域。


79. Phase M4 — Problem Identity / Theory Extension

加入:

problem relation classifier
problem mutation receipt
theory extension classifier
assumption-envelope update

成功條件:

SilentMutationRate.SilentMutationRate \downarrow.

80. Phase M5 — Obligation Engine

rule-first obligation templates。

成功條件:

RequiredObligationRecall>Baseline.RequiredObligationRecall > Baseline.

同時控制 false obligations。


81. Phase M6 — Math Cognitive Registry

先建立 15–25 個 operator。

可:

resolve
validate
render
compose
replay

82. Phase M7 — Verifier Router

第一版只接少量 verifier class:

formal proof
symbolic
numeric / exhaustive counterexample
semantic review

先證明 routing semantics,再擴 provider。


83. Phase M8 — CMDC Repair Engine

支援:

definition repair
assumption candidate
representation change
lemma candidate
method retrieval

所有 repair 只建立 candidate state。


84. Phase M9 — Theory Bridge / Prior Art

加入:

known concept mapping
formal-library refs
literature refs
bridge strength
novelty status

85. Phase M10 — CTCL / Receipt Adapter

所有重大 math events:

TemporalEvent.\rightarrow TemporalEvent.

並生成:

DiagnosisReceipt
RepairReceipt
ProblemMutationReceipt
TheoryExtensionReceipt
VerificationReceipt
ResearchReceipt

86. Phase M11 — Persistent Research Loop

最後才接:

ResearchGoal+Environment+Contract.ResearchGoal + Environment + Contract.

不再逐輪提供:

下一步請做什麼。


87. MVP 第一個環境

最適合不是直接丟 Millennium Problem。

第一個環境應是:

ControlledMathematicalResearchSandbox.\boxed{ ControlledMathematicalResearchSandbox. }

內含:

small theorem set
known malformed problems
missing assumptions
ambiguous definitions
false conjectures
missing lemmas
simple Lean targets
finite counterexample domains
known prior-art mappings

88. 第一個 Demo

Research Goal:

Investigate whether the target statement is valid and produce a reviewable result package.

Environment:

ambiguous definition
one hidden counterexample
formal target available
small prior-art corpus

期待:

Observe
→ DefinitionGap
→ ClarificationCandidate
→ ProblemRelation
→ ObligationGeneration
→ CounterexampleSearch
→ conjecture refuted
→ restricted branch proposed
→ verification
→ TheoryPackage
→ STOP / ESCALATE

89. 第二個 Demo — Silent Assumption Trap

原問題:

Q.Q.

模型容易加:

A.A.

期待:

T+AQT+A\vdash Q

不能被包成:

TQ.T\vdash Q.

測:

NoLaunderingGate.NoLaunderingGate.

90. 第三個 Demo — Wrong Domain

真 gap:

DefinitionGap.DefinitionGap.

比較:

MoreComputeMoreCompute

與:

DefinitionRepair.DefinitionRepair.

測:

MatchedRepairGain.MatchedRepairGain.

91. 第四個 Demo — False Conjecture

AI 不應無限 proof search。

期待:

CounterexampleReDiagnoseProblemBranch.Counterexample \rightarrow ReDiagnose \rightarrow ProblemBranch.

92. 第五個 Demo — Faithfulness Trap

自然語言問題:

QN.Q_N.

提供一個容易證但偏離原意的:

QF.Q_F.

測:

SemanticFaithfulnessGate.SemanticFaithfulnessGate.

93. 第六個 Demo — Prior Art

AI 生成新 definition。

但 corpus 中已有等價概念。

期待:

NovelCandidateRediscovered/EquivalentTo.NovelCandidate \rightarrow Rediscovered/EquivalentTo.

而不是:

NewTheory.NewTheory.

94. Falsification Gate A — State Determinism

EquivalentPublicResearchStateSameCanonicalState.EquivalentPublicResearchState \rightarrow SameCanonicalState.

95. Gate B — Gap Diagnosis

PredictedGapGroundTruthGap.PredictedGap \approx GroundTruthGap.

測 multi-label confusion matrix。


96. Gate C — Matched Affordance

MatchedMathOperator>RandomOperator.MatchedMathOperator > RandomOperator.

97. Gate D — Problem Identity

SilentMutationRate0.SilentMutationRate \rightarrow 0.

98. Gate E — Obligation Coverage

GeneratedObligationsRequiredObligations.GeneratedObligations \approx RequiredObligations.

99. Gate F — No Laundering

ClaimScopeVerifiedScope.ClaimScope \not> VerifiedScope.

100. Gate G — Verifier Routing

MatchedVerifier>SingleVerifierBaseline.MatchedVerifier > SingleVerifierBaseline.

101. Gate H — Repair Validity

RepairValidity=GapClosureUndeclaredDamage.RepairValidity = GapClosure - UndeclaredDamage.

102. Gate I — Gap Reopen

加入 delayed counterexample 後:

ClosedGapReopened?ClosedGap \rightarrow Reopened?

103. Gate J — Context Recovery

context 壓縮後能否用:

Ledger+ReceiptsLedger + Receipts

恢復:

problem lineage
assumption envelope
open obligations
verification basis

104. Gate K — Long-Horizon Research

只給:

ResearchGoal+Environment+ContractResearchGoal + Environment + Contract

測:

5010050\sim100

個 research transitions。


105. 評估向量

AMRR evaluator:

Score=(P,D,R,I,O,V,B,N,C,E).Score = ( P,D,R,I,O,V,B,N,C,E ).

其中:

  • PP:problem-solving performance;
  • DD:diagnosis;
  • RR:repair validity;
  • II:identity preservation;
  • OO:obligation quality;
  • VV:verification;
  • BB:bridge quality;
  • NN:novelty calibration;
  • CC:continuity / auditability;
  • EE:efficiency。

106. 不用單一 Reward

避免:

ScoreRScore\in\mathbb R

把:

CorrectAnswerCorrectAnswer

抵銷:

IdentityViolation.IdentityViolation.

因此初期使用 vector evaluation。


107. 成功與失敗的新定義

一個 final theorem 沒有證成,不一定是 research failure。

例如:

OriginalConjecture=FalseOriginalConjecture=False

且 AI 找到可驗證反例,這可能是:

ResearchSuccess.ResearchSuccess.

反過來:

ProofFoundProofFound

但偷加假設,則可能是:

ResearchGovernanceFailure.ResearchGovernanceFailure.

108. AMRR v0.1 不做什麼

第一版不做:

fully autonomous publication
unbounded axiom creation
unsupervised foundational replacement
claiming universal novelty
claiming universal mathematical correctness
philosophical consciousness
legal mathematical personhood
full automatic community acceptance

109. AMRR v0.1 真正只做什麼

第一版只回答:

給定一個受控數學研究環境、持續 Research Goal 與 Research Contract,AI 是否能在不需要人類逐輪指定下一步的情況下,持續診斷問題域、選擇數學認知程序、提出受約束 repair、保留問題身份、管理義務、調用 verifier、形成 theory state,並留下可恢復的研究因果歷史?


110. 與前三篇的統一

Paper 01:

AutonomousMathematicalResearch.AutonomousMathematicalResearch.

Paper 02:

DomainDiagnosis+CMDC.DomainDiagnosis + CMDC.

Paper 03:

ProblemIdentity+TheoryExtension+Obligations.ProblemIdentity + TheoryExtension + Obligations.

Paper 04:

ExecutableAMRR.\boxed{ ExecutableAMRR. }

111. 系列最終總模型

ResearchGoaltObserveMathematicalStatetDiagnoseGapMaptAgendatCognitiveProgramtRepair/ResearchCandidatetProblemIdentitytTheoryExtensiontObligationSettVerificationtGovernancetAcceptedTheoryStatet+1Bridge/NoveltytReceipts/LedgertReDiagnose.\boxed{ \begin{aligned} ResearchGoal_t &\rightarrow Observe\\ &\rightarrow MathematicalState_t\\ &\rightarrow Diagnose\\ &\rightarrow GapMap_t\\ &\rightarrow Agenda_t\\ &\rightarrow CognitiveProgram_t\\ &\rightarrow Repair/ResearchCandidate_t\\ &\rightarrow ProblemIdentity_t\\ &\rightarrow TheoryExtension_t\\ &\rightarrow ObligationSet_t\\ &\rightarrow Verification_t\\ &\rightarrow Governance_t\\ &\rightarrow AcceptedTheoryState_{t+1}\\ &\rightarrow Bridge/Novelty_t\\ &\rightarrow Receipts/Ledger_t\\ &\rightarrow ReDiagnose. \end{aligned} }

112. 系列最終不變量

UnsolvedSingleFailureMode.\boxed{ Unsolved \neq SingleFailureMode. } DiagnosisTruth.\boxed{ Diagnosis \neq Truth. } RepairCandidateAcceptedMutation.\boxed{ RepairCandidate \neq AcceptedMutation. } GenealogicalContinuity⇏SemanticEquivalence.\boxed{ GenealogicalContinuity \not\Rightarrow SemanticEquivalence. } T+AQ⇏TQ.\boxed{ T+A\vdash Q \not\Rightarrow T\vdash Q. } ProofFoundClaimApproved.\boxed{ ProofFound \neq ClaimApproved. } WAIVEDDISCHARGED.\boxed{ WAIVED \neq DISCHARGED. } ClaimScopeVerifiedScope.\boxed{ ClaimScope \not> VerifiedScope. } WorkingContextResearchLedger.\boxed{ WorkingContext \neq ResearchLedger. } SubmissionReady⇏CommunityAccepted.\boxed{ SubmissionReady \not\Rightarrow CommunityAccepted. }

113. 核心命題

本文最終提出:

 Autonomous mathematics should be implemented as a governed mathematical-state runtime, not as an unbounded loop of generated text. \boxed{ \textbf{ Autonomous mathematics should be implemented as a governed mathematical-state runtime, not as an unbounded loop of generated text. } }

中文:

自主數學不應被實作成一個不停產生文字與證明的 Agent,而應被實作成一個具有 canonical 數學狀態、問題與理論版本、typed gap、受約束 repair、數學義務、外部 verifier、治理決策與時間因果研究歷史的持續 Runtime。

更完整地:

AutonomousMathematics=SelfDirectedCognition+ExplicitMathematicalState+ConstrainedTheoryMutation+ExternalVerification+TemporalCausalAccountability.\boxed{ AutonomousMathematics = SelfDirectedCognition + ExplicitMathematicalState + ConstrainedTheoryMutation + ExternalVerification + TemporalCausalAccountability. }

114. 結論

數學 AI 若只會等待人類提供完整問題,它仍主要是一個 solver。

若它可以自己選 proof strategy,但不能知道自己其實缺的是定義、假設、表示或 theory bridge,它仍缺少研究層的 diagnosis。

若它可以自己改問題與建理論,卻不能保存 Problem Identity、Theory Extension 與 Assumption Envelope,它可能只是在「改到可解」。

若它可以產生 proof,卻沒有 typed verifier、semantic faithfulness 與 claim-scope gate,它可能只是在更可靠地證明錯誤形式化。

因此真正的自主數學研究需要:

State+Diagnosis+Cognition+Mutation+Obligation+Verification+Governance+History.\boxed{ State + Diagnosis + Cognition + Mutation + Obligation + Verification + Governance + History. }

本文提出 AMRR v0.1 作為這些能力的統一 Runtime 架構。

它不要求 AI 一開始就具有無界的數學創造力。

相反,它要求每一項自治能力都通過:

ArchitectureCapabilityFalsificationGate.\boxed{ Architecture \rightarrow Capability \rightarrow FalsificationGate. }

當這條路線完成後,人類與數學 AI 的接口才可能真正從:

HumanProblemAIAnswerHuman \rightarrow Problem \rightarrow AI \rightarrow Answer

逐步轉為:

HumanResearchContractAutonomousMathematicalResearchRuntime.\boxed{ Human \xleftrightarrow{ResearchContract} AutonomousMathematicalResearchRuntime. }

人類不再必須逐輪替 AI 撰寫每一個下一步;但數學身份、證明義務、理論依賴、權限邊界與研究歷史仍保持可檢查、可回退、可驗證。

這才是本系列所稱的:

從數學解題 AI 到自主數學研究 AI。\boxed{ \text{從數學解題 AI 到自主數學研究 AI。} }

參考文獻

[1] Feng, T., Trinh, T. H., Bingham, G., et al. (2026). Towards Autonomous Mathematics Research. arXiv:2602.10177.

[2] Tsoukalas, G., Kovsharov, A., Shirobokov, S., et al. (2026). Advancing Mathematics Research with AI-Driven Formal Proof Search. arXiv:2605.22763.

[3] Min, M. J., He, M., Li, Z., Yi, Z., Malik, S., Gupta, A., Si, X., & Bastani, O. (2026). Theory-Level Autoformalization: From Isolated Statements to Unified Formal Knowledge Bases. arXiv:2607.13292.

[4] Zhang, L., Valentino, M., & Freitas, A. (2025). Autoformalization in the Wild: Assessing LLMs on Real-World Mathematical Definitions. Proceedings of EMNLP 2025, 1720–1738. DOI: 10.18653/v1/2025.emnlp-main.90.

[5] Poiroux, A., Weiss, G., Kunčak, V., & Bosselut, A. (2025). Reliable Evaluation and Benchmarks for Statement Autoformalization. Proceedings of EMNLP 2025, 17947–17969. DOI: 10.18653/v1/2025.emnlp-main.907.

[6] Rabe, F., & Kohlhase, M. (2013). A Scalable Module System. Information and Computation, 230, 1–54. DOI: 10.1016/j.ic.2013.06.001.

[7] Xue, T. (2014). Definitional Extension in Type Theory. TYPES 2013, LIPIcs 26, 251–269. DOI: 10.4230/LIPIcs.TYPES.2013.251.

[8] Zhang, J., & Tan, S.-C. (2026). Automated Conjecturing and Theorem Finding: A Survey. Journal of Computer Science and Technology, 41(1), 46–66. DOI: 10.1007/s11390-026-6040-0.


內部架構依賴文件

[I1] Neo.K. (2026). 從數學解題到自主數學研究:受約束數學域補全與自主數學研究 Runtime,Autonomous Mathematical Research / Paper 01 of 04,v0.1.

[I2] Neo.K. (2026). 數學問題不是只有可解與不可解:多域問題診斷與受約束數學域補全,Autonomous Mathematical Research / Paper 02 of 04,v0.1.

[I3] Neo.K. (2026). 問題身份、理論擴張與數學義務:AI 生成數學的合法變換框架,Autonomous Mathematical Research / Paper 03 of 04,v0.1.

[I4] Neo.K. (2026). 從自提示到自主認知閉環:持續目標型 AI 的基礎理論,系列 01/06,v0.1.

[I5] Neo.K. (2026). 可定址認知空間:Cognitive Affordance、Semantic Address 與認知算子,系列 02/06,v0.1.

[I6] Neo.K. (2026). 自我對話不是文字:AI-Native Cognitive Program 與 Zero-Rendering Runtime,系列 03/06,v0.1.

[I7] Neo.K. (2026). 時間因果自我史:CTCL-ITR、Decision Receipt 與上下文壓縮後的可追溯性,系列 04/06,v0.1.

[I8] Neo.K. (2026). 契約邊界內的 AI 自主性:Execute、Refuse、Defer、Idle 與 Escalate,系列 05/06,v0.1.

[I9] Neo.K. (2026). Addressable Cognitive Runtime × CTCL:統一技術白皮書與實作路線圖 v0.1,系列 06/06.

[I10] Addressable Cognitive Runtime MVP v0.1.2. Phase 2 Validation: Semantic State Encoder. Internal implementation artifact, 2026-08-21.


版本備註

v0.1 / 2026-08-23

本版正式固定:

  1. AMRR 五層 Runtime architecture;
  2. extensions.math 兼容 ACR SemanticState 的策略;
  3. Problem Store / Problem DAG;
  4. Theory Store / Theory Extension Classifier;
  5. Gap Diagnosis Engine / Gap DAG;
  6. CMDC Repair Engine / MathChangeSet;
  7. Mathematical Obligation Engine / Obligation Graph;
  8. cog://math/* Cognitive Registry;
  9. Mathematical Research Program Compiler;
  10. Verifier Router / Verifier Ensemble;
  11. Assumption Envelope / No-Laundering Gate;
  12. Theory Bridge / Novelty Layer;
  13. Research Contract / Agenda / Governance;
  14. CTCL math event namespace 與 mathematical receipts;
  15. Persistent Research Loop;
  16. Phase M0–M11 實作順序;
  17. Controlled Mathematical Research Sandbox;
  18. Gate A–K falsification program;
  19. vector evaluation;
  20. 與 ACR Phase 0–2 既有工程的相容性邊界。

本版仍是 architecture specification paper,不宣稱 AMRR 的 M0–M11 已全部實作。現有工程基線只採既有 ACR Phase 0–2;所有新增 math runtime components 應在後續 technical whitepaper / MVP 中逐項實作與否證。