Sync GitHub main af692537c27175c8410bf57a747be0caf23718e2

MIRROR_MANIFEST.json

@@ -1,6 +1,6 @@
 {
   "canonical_repository": "https://github.com/Thomas-Heisig/MHRN",
-  "canonical_sha": "e8a04aea24121d15707238534e635cf0ceee8eaf",
+  "canonical_sha": "af692537c27175c8410bf57a747be0caf23718e2",
   "generated_by": ".github/workflows/sync-huggingface.yml",
   "rolling_mirror": true,
   "raw_experiment_data_included": false,

research/RESEARCH_CATALOG.json

@@ -4,12 +4,12 @@
   "scope": "research",
   "warning": "Classification describes document role; it does not promote DATA to EVID or certify scientific validity.",
   "summary": {
-    "file_count": 8774,
+    "file_count": 8786,
     "status": {
       "archive": 2,
-      "current": 233,
-      "current_wip": 41,
-      "experimental": 7962,
+      "current": 242,
+      "current_wip": 42,
+      "experimental": 7964,
       "frozen": 83,
       "generated": 63,
       "historical": 383,
@@ -21,8 +21,9 @@
       "archive_metadata_index": 1,
       "benchmark": 2,
       "critique": 1,
+      "engineering_verification_manifest": 2,
       "ethics_policy": 4,
-      "experiment": 7942,
+      "experiment": 7944,
       "experiment_execution_receipt": 1,
       "experiment_workflow": 13,
       "frontier_programme": 1,
@@ -33,12 +34,12 @@
       "learning_proposal_revision": 1,
       "literature": 11,
       "method_calibration": 5,
-      "preregistration": 92,
+      "preregistration": 93,
       "prior_art_review_programme": 1,
-      "protocol": 17,
+      "protocol": 18,
       "publication": 507,
       "publication_pointer": 2,
-      "registry": 52,
+      "registry": 55,
       "related_work_register": 1,
       "report_generator": 1,
       "research_crosswalk_configuration": 1,
@@ -55,17 +56,17 @@
       "scientific_data_contract": 1,
       "scientific_execution_audit": 1,
       "scientific_execution_recovery_record": 1,
-      "scientific_governance_decision": 2,
+      "scientific_governance_decision": 4,
       "scientific_method_decision": 4,
       "scientific_state": 1,
-      "specification": 8,
+      "specification": 9,
       "stage_research_summary": 2
     },
     "authority": {
       "accepted_architecture_and_research_scope_decision": 1,
       "accepted_execution_provenance_record": 2,
       "accepted_implementation_and_method_decision": 1,
-      "accepted_method_and_governance_decision": 2,
+      "accepted_method_and_governance_decision": 4,
       "accepted_scientific_method_decision": 2,
       "accepted_stage1_scientific_consolidation_decision": 1,
       "analysis_method_configuration": 1,
@@ -73,28 +74,30 @@
       "archive_provenance_index": 1,
       "benchmark_reference_or_method": 2,
       "canonical_current_publication_pointer": 1,
+      "canonical_frozen_environment_verification_input": 1,
+      "canonical_pan_backend_verification_input": 1,
       "canonical_research_navigation": 1,
       "critical_analysis": 1,
       "current_scientific_state_summary": 1,
       "derived_report_tooling": 1,
       "descriptive_taxonomy_coordinates": 1,
       "execution_record": 1,
-      "experiment_artifact": 7942,
-      "frozen_or_versioned_experiment_plan": 92,
+      "experiment_artifact": 7944,
+      "frozen_or_versioned_experiment_plan": 93,
       "generated_projection": 63,
       "historical_context": 1,
       "historical_empirical_baseline": 82,
       "historical_integrative_predecessor": 7,
       "human_or_external_review_record": 4,
-      "normative_design_contract": 8,
+      "normative_design_contract": 9,
       "normative_document_governance_configuration": 1,
       "normative_ethics_context": 4,
       "normative_meta_system_configuration": 1,
       "normative_meta_system_integration_contract": 1,
-      "normative_protocol_or_method": 17,
+      "normative_protocol_or_method": 18,
       "normative_research_development_policy": 1,
       "normative_research_integrity_policy": 1,
-      "normative_research_registry": 52,
+      "normative_research_registry": 55,
       "normative_schema": 9,
       "normative_seed_provenance_specification": 1,
       "pre_execution_method_calibration": 5,
@@ -483,6 +486,32 @@
       "classification_rule": "override:research/decisions",
       "review_required": false
     },
+    {
+      "path": "research/decisions/2026-09-29_playground_mhrn_integration_wave1.md",
+      "domain": "research",
+      "kind": "scientific_governance_decision",
+      "status": "current",
+      "authority": "accepted_method_and_governance_decision",
+      "mutability": "versioned_replacement_only",
+      "citation": "cite_decision_path_revision_and_underlying_sources",
+      "evidence_role": "decision_or_interpretation_not_evidence_by_itself",
+      "rationale": "Scientific decisions preserve corrections, stop rules and claim boundaries without altering historical DATA.",
+      "classification_rule": "override:research/decisions",
+      "review_required": false
+    },
+    {
+      "path": "research/decisions/2026-10-01_wave6_first_governed_run_mode.md",
+      "domain": "research",
+      "kind": "scientific_governance_decision",
+      "status": "current",
+      "authority": "accepted_method_and_governance_decision",
+      "mutability": "versioned_replacement_only",
+      "citation": "cite_decision_path_revision_and_underlying_sources",
+      "evidence_role": "decision_or_interpretation_not_evidence_by_itself",
+      "rationale": "Scientific decisions preserve corrections, stop rules and claim boundaries without altering historical DATA.",
+      "classification_rule": "override:research/decisions",
+      "review_required": false
+    },
     {
       "path": "research/document_governance_overrides.json",
       "domain": "research",
@@ -92133,6 +92162,32 @@
       "classification_rule": "override:research/experiments",
       "review_required": false
     },
+    {
+      "path": "research/experiments/EXP-GEN-0048/manifest.json",
+      "domain": "research",
+      "kind": "experiment",
+      "status": "experimental",
+      "authority": "experiment_artifact",
+      "mutability": "append_only_or_immutable_after_completion",
+      "citation": "cite_experiment_id_path_digest",
+      "evidence_role": "data_or_review_state_as_declared_by_manifest",
+      "rationale": "Experiment directories preserve execution provenance; completion does not imply accepted EVID.",
+      "classification_rule": "override:research/experiments",
+      "review_required": false
+    },
+    {
+      "path": "research/experiments/EXP-GEN-0048/workflow.json",
+      "domain": "research",
+      "kind": "experiment",
+      "status": "experimental",
+      "authority": "experiment_artifact",
+      "mutability": "append_only_or_immutable_after_completion",
+      "citation": "cite_experiment_id_path_digest",
+      "evidence_role": "data_or_review_state_as_declared_by_manifest",
+      "rationale": "Experiment directories preserve execution provenance; completion does not imply accepted EVID.",
+      "classification_rule": "override:research/experiments",
+      "review_required": false
+    },
     {
       "path": "research/experiments/EXP-LEARN-20260903-2300/DATA/runs.json",
       "domain": "research",
@@ -104899,6 +104954,19 @@
       "classification_rule": "research-zone:preregistrations",
       "review_required": false
     },
+    {
+      "path": "research/preregistrations/PREREG-CPU-PAR-001.json",
+      "domain": "research",
+      "kind": "preregistration",
+      "status": "current",
+      "authority": "frozen_or_versioned_experiment_plan",
+      "mutability": "immutable_after_registration_or_versioned_replacement",
+      "citation": "cite_preregistration_id_and_revision",
+      "evidence_role": "preregistration_not_result",
+      "rationale": "Rule-classified repository documentation; exact scientific authority depends on referenced DATA/EVID.",
+      "classification_rule": "research-zone:preregistrations",
+      "review_required": false
+    },
     {
       "path": "research/preregistrations/PREREG-EMB001B-MATCHED-V2.json",
       "domain": "research",
@@ -106121,6 +106189,19 @@
       "classification_rule": "research-zone:protocols",
       "review_required": false
     },
+    {
+      "path": "research/protocols/CPU_SELF_PARITY_V1.md",
+      "domain": "research",
+      "kind": "protocol",
+      "status": "current",
+      "authority": "normative_protocol_or_method",
+      "mutability": "editable",
+      "citation": "cite_exact_path_and_revision",
+      "evidence_role": "protocol_not_result",
+      "rationale": "Rule-classified repository documentation; exact scientific authority depends on referenced DATA/EVID.",
+      "classification_rule": "research-zone:protocols",
+      "review_required": false
+    },
     {
       "path": "research/protocols/EXP_GEN_0021_FOLLOWUP_PROGRAM.yaml",
       "domain": "research",
@@ -113375,6 +113456,19 @@
       "classification_rule": "override:research/registry",
       "review_required": false
     },
+    {
+      "path": "research/registry/hypotheses.acceleration.yaml",
+      "domain": "research",
+      "kind": "registry",
+      "status": "current",
+      "authority": "normative_research_registry",
+      "mutability": "editable_with_review",
+      "citation": "cite_exact_object_and_revision",
+      "evidence_role": "research_definition_not_evidence_by_itself",
+      "rationale": "Canonical RQ/H/claim/experiment registry definitions.",
+      "classification_rule": "override:research/registry",
+      "review_required": false
+    },
     {
       "path": "research/registry/hypotheses.behavior_memory.yaml",
       "domain": "research",
@@ -113531,6 +113625,19 @@
       "classification_rule": "override:research/registry",
       "review_required": false
     },
+    {
+      "path": "research/registry/questions.acceleration.yaml",
+      "domain": "research",
+      "kind": "registry",
+      "status": "current",
+      "authority": "normative_research_registry",
+      "mutability": "editable_with_review",
+      "citation": "cite_exact_object_and_revision",
+      "evidence_role": "research_definition_not_evidence_by_itself",
+      "rationale": "Canonical RQ/H/claim/experiment registry definitions.",
+      "classification_rule": "override:research/registry",
+      "review_required": false
+    },
     {
       "path": "research/registry/questions.behavior_memory.yaml",
       "domain": "research",
@@ -113648,6 +113755,19 @@
       "classification_rule": "override:research/registry",
       "review_required": false
     },
+    {
+      "path": "research/registry/sources.acceleration.yaml",
+      "domain": "research",
+      "kind": "registry",
+      "status": "current",
+      "authority": "normative_research_registry",
+      "mutability": "editable_with_review",
+      "citation": "cite_exact_object_and_revision",
+      "evidence_role": "research_definition_not_evidence_by_itself",
+      "rationale": "Canonical RQ/H/claim/experiment registry definitions.",
+      "classification_rule": "override:research/registry",
+      "review_required": false
+    },
     {
       "path": "research/registry/sources.cognition.yaml",
       "domain": "research",
@@ -113986,6 +114106,19 @@
       "classification_rule": "research-zone:specifications",
       "review_required": false
     },
+    {
+      "path": "research/specifications/WAVE7_LEARNING_DIVERGENCES.json",
+      "domain": "research",
+      "kind": "specification",
+      "status": "current",
+      "authority": "normative_design_contract",
+      "mutability": "editable",
+      "citation": "cite_exact_path_and_revision",
+      "evidence_role": "contract_not_evidence",
+      "rationale": "Rule-classified repository documentation; exact scientific authority depends on referenced DATA/EVID.",
+      "classification_rule": "research-zone:specifications",
+      "review_required": false
+    },
     {
       "path": "research/stage4_specialized_neural_areas.md",
       "domain": "research",
@@ -114012,6 +114145,32 @@
       "classification_rule": "override:research/stage5_integrated_nervous_system.md",
       "review_required": false
     },
+    {
+      "path": "research/verification/frozen_environment/FE3_DETERMINISTIC_TARGET_V1.json",
+      "domain": "research",
+      "kind": "engineering_verification_manifest",
+      "status": "current",
+      "authority": "canonical_frozen_environment_verification_input",
+      "mutability": "versioned_replacement_only",
+      "citation": "cite_exact_path_manifest_sha256_and_source_revision",
+      "evidence_role": "engineering_verification_input_not_data_or_evidence",
+      "rationale": "Versioned deterministic Frozen-Environment verification inputs used for backend parity acceptance; these manifests are engineering verification inputs and do not constitute scientific DATA or EVID.",
+      "classification_rule": "override:research/verification/frozen_environment",
+      "review_required": false
+    },
+    {
+      "path": "research/verification/pan/PAN_WAVE5B_PREFLIGHT_V1.json",
+      "domain": "research",
+      "kind": "engineering_verification_manifest",
+      "status": "current_wip",
+      "authority": "canonical_pan_backend_verification_input",
+      "mutability": "versioned_replacement_only",
+      "citation": "cite_exact_manifest_path_digest_and_source_revision",
+      "evidence_role": "engineering_verification_input_not_data_or_evidence",
+      "rationale": "Versioned PAN backend parity preflight inputs and future authorization records; preparation and execution receipts do not automatically constitute scientific DATA or EVID.",
+      "classification_rule": "override:research/verification/pan",
+      "review_required": false
+    },
     {
       "path": "research/workflows/EXP-BATCH-20260906200118-recovery.md",
       "domain": "research",

research/decisions/2026-09-29_playground_mhrn_integration_wave1.md

@@ -0,0 +1,66 @@
+# Decision: Playground -> MHRN Integration Wave 1
+
+**Date:** 2026-09-29  
+**Status:** accepted engineering integration decision  
+**Evidence status:** engineering provenance only; no DATA/EVID promotion
+
+## Decision
+
+Playground components are promoted into canonical MHRN one contract family at a
+time. The dependency direction is fixed:
+
+```text
+Playground -> canonical MHRN
+canonical MHRN -X-> Playground
+```
+
+The first promoted family is the typed neural-I/O boundary contract:
+
+- `BoundaryFrame`
+- `CodecContract`
+- `PopulationLayout`
+- `SpikeEvent` / `SpikeFrame`
+- `DecodeResult`
+- `NeuralRole` / `InterfacePhase`
+- deterministic payload/event/readout digest helpers
+
+Canonical ownership is now
+`src/embodiment/neural_io_contracts.py`. The historical
+`src/playground/neural_io/contracts.py` path remains as a compatibility
+re-export so existing Playground codecs and sessions keep the same class
+identity.
+
+## Transfer surface
+
+The Playground Dashboard exposes the integration state through:
+
+- `GET /api/playground/integration`
+- `POST /api/playground/integration/transfer`
+
+The transfer action is deliberately **not** a runtime source-code editor. It
+verifies already integrated elements or reports the next required gate for a
+future reviewed repository change.
+
+## OLD frontend boundary
+
+The OLD workspace remains a compatibility/archive surface. Views stored there
+may be referenced in the integration catalog, but they are never promoted
+implicitly into MHRN core merely because they remain executable.
+
+## Promotion order after wave 1
+
+1. deterministic neural-I/O codecs after codec/frame-ID semantics freeze;
+2. CUDA Driver/NVRTC/ABI/parity behind a canonical ExecutionBackend;
+3. learning/plasticity only after the canonical Learning/Synapse contract;
+4. closed-loop/environment only after FE-1/FE-2/FE-3 Frozen-Environment gates;
+5. PAN hyperstate only after RQ-PAN-SEM-001 semantic freeze.
+
+## Scientific boundary
+
+This decision changes software ownership and interface provenance only. It does
+not create a new experiment, scientific DATA, EVID, replication, cognition
+claim, PAN validity claim, or CPU/CUDA scientific-equivalence claim.
+
+Future scientific work may use the promoted neural-I/O contract as a shared
+measurement boundary, but any resulting claim still requires preregistration,
+fresh DATA, controls and Human Review under the existing research governance.

research/decisions/2026-10-01_wave6_first_governed_run_mode.md

@@ -0,0 +1,39 @@
+# Governance Decision: Wave 6 first governed structural run
+
+**Decision ID:** `DEC-WAVE6-FIRST-RUN-MANUAL-20261001`  
+**Date:** 2026-10-01  
+**Status:** accepted governance selection; execution not authorized  
+**Contract:** `mhrn-structural-approval-v0.1`
+
+## Decision
+
+The first future governed structural-plasticity run will use `ApprovalMode.MANUAL_ONLY`.
+
+The system default remains `DISABLED`. This decision does not enable structural mutation and does not satisfy the independent hard gates by itself.
+
+## Rationale
+
+`MANUAL_ONLY` is selected for the first governed run because it permits direct inspection of proposals and journal/undo behavior without introducing an unreviewed automatic approval policy.
+
+`POLICY_AUTO` remains deferred until the policy has its own validation. `PREREGISTERED_AUTO` remains deferred until an exact policy artifact is frozen inside a preregistered structural study.
+
+## Preconditions before execution authorization
+
+1. exact structural config is frozen;
+2. policy artifact hash is recorded;
+3. manual authorization is explicitly recorded;
+4. topology changes are permitted by the study/protocol;
+5. structural barrier is available;
+6. journal and undo path are healthy;
+7. scientific freeze permits the mutation;
+8. normal safety/resource/cooldown/kind checks pass.
+
+No CUDA in-kernel structural mutation is authorized by this decision.
+
+## Separation from first CPU preregistration
+
+`PREREG-CPU-PAR-001` is intentionally executed with structural approval `DISABLED`. The self-parity study must not conflate CPU determinism with topology adaptation.
+
+## Evidence boundary
+
+This is a governance/method decision, not DATA, EVID or a structural-plasticity effect claim.

research/document_governance_overrides.json

@@ -461,6 +461,105 @@
       "citation": "cite_exact_path_and_revision_as_playground_context",
       "evidence_role": "not_evidence_by_itself",
       "rationale": "Documents the explicitly non-canonical Playground surface. Playground documentation, presets and exploratory runtime descriptions do not create DATA or EVID."
+    },
+    "docs/canonical/EXECUTION_BACKEND_CONTRACT.md": {
+      "kind": "canonical_architecture_contract",
+      "status": "current",
+      "authority": "normative_execution_backend_contract",
+      "mutability": "editable_with_architecture_review",
+      "citation": "cite_current_revision_and_implementation_commit",
+      "evidence_role": "engineering_contract_not_evidence",
+      "rationale": "Defines the backend-neutral ExecutionBackend interface and state/capability boundary introduced by Playground integration Wave 3; it is an engineering contract and creates no DATA or EVID."
+    },
+    "docs/canonical/PARITY_CONTRACT.md": {
+      "kind": "canonical_verification_contract",
+      "status": "current",
+      "authority": "normative_cross_backend_parity_contract",
+      "mutability": "editable_with_architecture_and_method_review",
+      "citation": "cite_current_revision_and_underlying verification artifacts",
+      "evidence_role": "verification_method_contract_not_evidence",
+      "rationale": "Defines D1/D2/D3 parity, stable execution fingerprints and fail-closed comparison semantics; parity verification is not scientific evidence by itself."
+    },
+    "docs/canonical/README.md": {
+      "kind": "documentation_index",
+      "status": "current",
+      "authority": "canonical_contract_navigation",
+      "mutability": "editable_with_review",
+      "citation": "navigation_only_cite_underlying_contract",
+      "evidence_role": "navigation_not_evidence",
+      "rationale": "Index for current canonical MHRN architecture and verification contracts."
+    },
+    "docs/canonical/CUDA_BACKEND.md": {
+      "kind": "canonical_acceleration_contract",
+      "status": "current",
+      "authority": "normative_cuda_backend_contract",
+      "mutability": "editable_with_architecture_and_method_review",
+      "citation": "cite_current_revision_implementation_commit_and_hardware_verification_scope",
+      "evidence_role": "engineering_contract_not_evidence",
+      "rationale": "Defines the bounded canonical CUDA backend, kernel-freeze boundary, draft plasticity semantics and D1/D2 acceptance scope introduced by Playground integration Wave 4; it creates no DATA or EVID."
+    },
+    "docs/canonical/HARDWARE_ACCEPTANCE.md": {
+      "kind": "canonical_hardware_acceptance_contract",
+      "status": "current",
+      "authority": "normative_engineering_hardware_acceptance_contract",
+      "mutability": "editable_with_architecture_and_method_review",
+      "citation": "cite_current_revision_and_physical_hardware_artifact_when_available",
+      "evidence_role": "engineering_verification_contract_not_data_or_evidence",
+      "rationale": "Defines the bounded physical CUDA and Frozen-Environment FE-3 acceptance procedure, artifact format and interpretation limits; execution under this contract remains engineering verification and creates no DATA or EVID."
+    },
+    "research/verification/frozen_environment": {
+      "kind": "engineering_verification_manifest",
+      "status": "current",
+      "authority": "canonical_frozen_environment_verification_input",
+      "mutability": "versioned_replacement_only",
+      "citation": "cite_exact_path_manifest_sha256_and_source_revision",
+      "evidence_role": "engineering_verification_input_not_data_or_evidence",
+      "rationale": "Versioned deterministic Frozen-Environment verification inputs used for backend parity acceptance; these manifests are engineering verification inputs and do not constitute scientific DATA or EVID."
+    },
+    "docs/canonical/PAN_HYPERSTATE_CONTRACT_DRAFT.md": {
+      "kind": "canonical_semantic_contract_draft",
+      "status": "current_wip",
+      "authority": "normative_pan_semantic_contract_draft",
+      "mutability": "editable_until_explicit_freeze",
+      "citation": "cite_current_revision_as_draft_contract_only",
+      "evidence_role": "semantic_contract_draft_not_data_or_evidence",
+      "rationale": "Defines the backend-neutral PAN hyperstate reference semantics prepared in Wave 5A. The contract is explicitly draft, does not establish CPU/CUDA equivalence, and creates no DATA, EVID or accepted scientific claim."
+    },
+    "docs/canonical/PAN_WAVE5B_PREFLIGHT.md": {
+      "kind": "canonical_engineering_preflight_contract",
+      "status": "current_wip",
+      "authority": "normative_pan_wave5b_preflight_contract",
+      "mutability": "editable_until_wave5b_freeze",
+      "citation": "cite_current_revision_as_engineering_preflight_only",
+      "evidence_role": "engineering_preflight_not_data_or_evidence",
+      "rationale": "Defines the fail-closed Wave-5B PAN CPU/CUDA parity preparation, gate ordering and interpretation boundary without authorizing execution or promoting PAN to canonical integrated status."
+    },
+    "research/verification/pan": {
+      "kind": "engineering_verification_manifest",
+      "status": "current_wip",
+      "authority": "canonical_pan_backend_verification_input",
+      "mutability": "versioned_replacement_only",
+      "citation": "cite_exact_manifest_path_digest_and_source_revision",
+      "evidence_role": "engineering_verification_input_not_data_or_evidence",
+      "rationale": "Versioned PAN backend parity preflight inputs and future authorization records; preparation and execution receipts do not automatically constitute scientific DATA or EVID."
+    },
+    "docs/canonical/STRUCTURAL_APPROVAL_CONTRACT_DRAFT.md": {
+      "kind": "canonical_governance_contract_draft",
+      "status": "current_wip",
+      "authority": "normative_structural_approval_contract_draft",
+      "mutability": "editable_until_explicit_governance_freeze",
+      "citation": "cite_current_revision_as_draft_contract_only",
+      "evidence_role": "governance_contract_draft_not_data_or_evidence",
+      "rationale": "Defines Wave-6 fail-closed structural approval modes, policy provenance and host-barrier prerequisites around existing proposal/coordinator/journal/undo mechanisms. It is a reviewed draft contract and does not authorize structural mutation or create DATA/EVID."
+    },
+    "docs/canonical/LEARNING_CONTRACT_DRAFT.md": {
+      "kind": "canonical_semantic_contract_draft",
+      "status": "current_wip",
+      "authority": "normative_learning_synapse_contract_draft",
+      "mutability": "editable_until_explicit_semantic_freeze",
+      "citation": "cite_current_revision_as_draft_contract_only",
+      "evidence_role": "semantic_contract_draft_not_data_or_evidence",
+      "rationale": "Defines the Wave-7 inventory of CPU learning semantics and explicit CPU/CUDA divergence blockers. It is a reviewed draft, does not establish cross-backend learning equivalence and creates no DATA/EVID."
     }
   }
 }