{ "section": "systems", "kind": "system", "canonical": "https://evemisslab.com/ai/systems/", "count": 7, "records": [ { "id": "SYS-2026-0002", "kind": "system", "label": "PACC-Lab — micro-lab harness for the convergence conjecture", "created_at": "2026-09-08", "updated_at": "2026-09-09", "values": { "eml_status": "EXPERIMENTAL", "eml_evidence_level": "E3", "eml_object_version": "0.1", "eml_canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0002/", "eml_provenance": { "source": "EveMissLab research collection: Adaptive Epistemic Systems (真本體論13)", "extracted_by": "Splice (Claude Code), reading the canonical UTF-8 sources and each lab's own result reports", "extracted_at": "2026-09-11", "generator": "tools/extract_aes/extract.py", "claim_boundary": "status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports" }, "eml_summary": "Fully observable evidence-integration environments, four primitive-level non-probabilistic families (N0–N3), exact Bayesian references, train-only low-complexity mappers, shuffled/constant/broken-composition controls, a design-redundancy diagnostic and frozen preregistered thresholds carried unchanged from v0.1 through v0.13. Each release is a sealed FINAL bundle with its protocol, results and next-experiment preregistration.", "eml_summary_zh": "可完全觀察的證據整合環境、四個 primitive 層非概率家族（N0–N3）、精確 Bayesian 參考、只在訓練集擬合的低複雜度映射器、shuffled／constant／broken-composition 控制組、設計冗餘診斷，以及從 v0.1 到 v0.13 原封不動的預登記門檻。每個版本都是密封的 FINAL 套件，含協定、結果與下一步實驗的預登記。", "eml_label_zh": "PACC-Lab——收斂猜想的微型實驗室", "eml_primary_domain": "Model Representation", "eml_domains": [ "Evaluation", "Formal AI" ], "eml_program_id": "PRG-2026-0001", "eml_purpose": "Test PACC-B/R/D/A with preregistered gates, negative controls and an explicit family-independence rule.", "eml_release": [ "PACC-Lab_v0.1_E0-E4_FINAL.zip", "PACC-Lab_v0.2_N3_Adversarial_FINAL.zip", "PACC-Lab_v0.3_Learned_Reliability_FINAL.zip", "PACC-Lab_v0.4_Correlated_Sources_FINAL.zip", "PACC-Lab_v0.5_Representation_Sufficiency_FINAL.zip", "PACC-Lab_v0.6_Composed_Coordinate_FINAL.zip", "PACC-Lab_v0.7_Cross_Geometry_Transfer_FINAL.zip", "PACC-Lab_v0.8_Transfer_Basin_FINAL.zip", "PACC-Lab_v0.9_Coordinate_Atlas_FINAL.zip", "PACC-Lab_v0.10_Cocycle_Coherence_FINAL.zip", "PACC-Lab_v0.11_Bidirectional_Groupoid_FINAL.zip", "PACC-Lab_v0.12_Bounded_Nonlinear_Transition_FINAL.zip", "PACC-Lab_v0.13_Residual_Field_FINAL.zip" ], "eml_authors": [ "Neo.K (EveMissLab)" ], "eml_ai_collaborators": [ "Sol (GPT-5.6, OpenAI ChatGPT)" ] }, "canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0002/", "json": "/ai/systems/SYS-2026-0002/index.json" }, { "id": "SYS-2026-0003", "kind": "system", "label": "PACC-LLM Hybrid Lab", "created_at": "2026-09-09", "updated_at": "2026-09-09", "values": { "eml_status": "EXPERIMENTAL", "eml_evidence_level": "E2", "eml_object_version": "0.1", "eml_canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0003/", "eml_provenance": { "source": "EveMissLab research collection: Adaptive Epistemic Systems (真本體論13)", "extracted_by": "Splice (Claude Code), reading the canonical UTF-8 sources and each lab's own result reports", "extracted_at": "2026-09-11", "generator": "tools/extract_aes/extract.py", "claim_boundary": "status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports" }, "eml_summary": "An A/B/C harness — generator only, hard verifier, PACC runtime — over shared candidate pools with equal accounted budgets. v0.1 is a synthetic architecture witness (zero LLM calls); v0.2 adds hand-authored natural-language tasks, a condition-blind judge, literal machine checks independent of the judge, fail-closed behaviour without an API key, and frozen-cache replay for exact reproduction after a live run.", "eml_summary_zh": "A/B/C harness——純生成器、硬驗證器、PACC runtime——在共享候選池與相同計費預算上運行。v0.1 是合成的架構見證（零次 LLM 呼叫）；v0.2 加入手寫自然語言任務、對條件盲的評審、獨立於評審的字面機器檢查、缺 API key 時 fail-closed，以及 live run 後可精確重播的凍結快取。", "eml_label_zh": "PACC-LLM 混合實驗室", "eml_primary_domain": "Reasoning", "eml_domains": [ "Evaluation" ], "eml_program_id": "PRG-2026-0001", "eml_purpose": "Decide whether a PACC-style commit space changes reasoning coherence, intent handling and creative breadth — first synthetically, then on a real language model.", "eml_release": [ "PACC-Hybrid-Lab_v0.1_SYNTHETIC_FINAL.zip", "PACC-Hybrid-Lab_v0.2_REAL_LLM_HARNESS_FINAL.zip" ], "eml_authors": [ "Neo.K (EveMissLab)" ], "eml_ai_collaborators": [ "Sol (GPT-5.6, OpenAI ChatGPT)" ] }, "canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0003/", "json": "/ai/systems/SYS-2026-0003/index.json" }, { "id": "SYS-2026-0001", "kind": "system", "label": "AER-0 — Adaptive Epistemic Runtime MVP", "created_at": "2026-09-08", "updated_at": "2026-09-08", "values": { "eml_status": "EXPERIMENTAL", "eml_evidence_level": "E2", "eml_object_version": "0.1", "eml_canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0001/", "eml_provenance": { "source": "EveMissLab research collection: Adaptive Epistemic Systems (真本體論13)", "extracted_by": "Splice (Claude Code), reading the canonical UTF-8 sources and each lab's own result reports", "extracted_at": "2026-09-11", "generator": "tools/extract_aes/extract.py", "claim_boundary": "status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports" }, "eml_summary": "A Python + SQLite reference runtime that owns persistent canonical state; a language model is optional and replaceable. Candidate → verify → versioned commit, mandatory provenance for facts and relations, freshness and confidence as separate state, an asymmetric tension scheduler with shock-triggered refresh, capability/container registries, persisted workflow memory (reuse/adapt/create), and a deterministic/rule/Bayesian/heuristic epistemic router. Grown through six comparison rounds into an epistemic commit transaction with SQLite-level mediation, an Ed25519-signed external anchor and a process-separated writer (79 tests at R6). Developed under mssp-tdd-apr as a DEGRADED-TWIN run — no independent twin verification is claimed.", "eml_summary_zh": "以 Python + SQLite 實作、自己持有持久 canonical state 的參考 runtime；語言模型可有可無、可替換。Candidate → verify → 版本化 commit、fact／relation 強制 provenance、新鮮度與信心分開存放、非對稱張力排程與 shock 觸發刷新、capability／container 註冊表、持久化 workflow 記憶（reuse／adapt／create）、以及 deterministic／rule／Bayesian／heuristic 的認識論 router。經六輪比較長成帶 SQLite 層中介、Ed25519 簽章外部 anchor 與程序分離 writer 的 epistemic commit transaction（R6 時 79 個測試）。在 mssp-tdd-apr 下以 DEGRADED-TWIN 開發——不主張獨立 twin 驗證。", "eml_label_zh": "AER-0——自適應認識 runtime MVP", "eml_primary_domain": "AI Architecture", "eml_domains": [ "Agent Systems", "AI Infrastructure" ], "eml_program_id": "PRG-2026-0001", "eml_purpose": "Make the Adaptive Epistemic Systems architecture executable and falsifiable — not to claim AGI.", "eml_architecture": "Input → Parse → Plan → Select capability/container → Execute → Verify → Candidate → Versioned Commit → Render → Learn workflow; refresh runs separately: canonical nodes → TensionScheduler → selected subset → refresh provider → verify → versioned commit. Only the Committer, after a passing Verifier verdict and expected-version check, may mutate canonical state.", "eml_release": [ "AER-0_MVP_v0.1_FINAL.zip", "AER-0_MVP_v0.1_Architecture_Comparison_R1.zip", "AER-0_MVP_v0.1_Architecture_Comparison_R2_LangGraph.zip", "AER-0_MVP_v0.1_Architecture_Comparison_R3_ECT.zip", "AER-0_MVP_v0.1_Architecture_Comparison_R4_Policy_Mutation.zip", "AER-0_MVP_v0.1_Architecture_Comparison_R5_Complete_Mediation.zip", "AER-0_MVP_v0.1_Architecture_Comparison_R6_External_Trust_Anchor.zip" ], "eml_known_limitations": [ "Not implemented: unrestricted generated-code execution, self-modifying runtime, distributed services, real quantum/biological backends, ontology self-rewrite, production authorization/deployment controls." ], "eml_authors": [ "Neo.K (EveMissLab)" ], "eml_ai_collaborators": [ "Sol (GPT-5.6, OpenAI ChatGPT)" ] }, "canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0001/", "json": "/ai/systems/SYS-2026-0001/index.json" }, { "id": "SYS-2026-0101", "kind": "system", "label": "XA-03 — telemetry and run logger (physical execution evidence layer)", "created_at": "2026-09-03", "updated_at": "2026-09-03", "values": { "eml_status": "STABLE", "eml_evidence_level": "E2", "eml_object_version": "0.1", "eml_canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0101/", "eml_provenance": { "source": "EveMissLab research collection: Intelligence Physical Metrology (真本體論13)", "extracted_by": "Splice (Claude Code), reading the canonical UTF-8 sources and each package's own reports", "extracted_at": "2026-09-11", "generator": "tools/extract_all.py", "claim_boundary": "status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports" }, "eml_summary": "Provider-agnostic logger that records one benchmark run as append-only events plus telemetry and derives a typed summary deterministically: model-invocation, trajectory, tool-call and verifier spans, candidate created/discarded accounting, wall time, device-time, GPU power integrated to device_energy_j (energy type device_measured, never relabelled as marginal), peak memory and memory residency, with unknowns kept null (Unknown ≠ 0) and forbidden conversions (tokens → J, TDP → J, price → J, GPU-hours → J). Golden fixture 450 J / 1.75 util·s / 12 GB peak / 33 GB·s residency verified; 35 passed, 1 skipped in 0.39s; instrumentation burden is calibrated and reported, not subtracted.", "eml_summary_zh": "與模型供應商無關的記錄器：把一次 benchmark 執行記成只增不改的事件流與遙測，再確定性地導出型別化摘要——模型呼叫、軌跡、工具呼叫與驗證器區段，候選建立／丟棄會計，wall time、裝置時間、GPU 功率積分成 device_energy_j（能量型別 device_measured，絕不改標成 marginal）、峰值記憶體與記憶體駐留，未知值保持 null（Unknown ≠ 0），並禁止 tokens → J、TDP → J、price → J、GPU-hours → J 的換算。黃金夾具 450 J／1.75 util·s／12 GB 峰值／33 GB·s 駐留驗證通過；35 passed, 1 skipped in 0.39s；儀器負擔經校準並報告，不自動扣除。", "eml_label_zh": "XA-03——遙測與執行記錄器（物理執行證據層）", "eml_primary_domain": "Computation", "eml_program_id": "PRG-2026-0101", "eml_purpose": "Make the physical side of an IPM event (wall time, device energy, memory residency, utilization, operational counts, discarded work) reproducible evidence rather than a claim.", "eml_architecture": "RunSession → EventLogger (paired spans) + TelemetryCollector (Null / System via psutil / NvidiaSmi) → MetricIntegrator (deterministic) → RunSink with atomic finalization (.partial → COMPLETE | ABORTED) and crash recovery that never invents end events; JSON schemas for events, telemetry and run summary.", "eml_documentation": "README.md, SPEC.md (36 sections), IMPLEMENTATION_PLAN.md, validation_report.json inside the package", "eml_tags": [ "EML-IPM-XA-03", "v0.1", "PRE_RELEASE_VERIFIED", "Python ≥3.11, no runtime dependencies" ], "eml_authors": [ "Neo.K (EveMissLab)" ], "eml_ai_collaborators": [ "Aletheia (GPT-5.6 Sol, OpenAI ChatGPT)" ] }, "canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0101/", "json": "/ai/systems/SYS-2026-0101/index.json" }, { "id": "SYS-2026-0102", "kind": "system", "label": "XA-04 — A0→A5 model runner and scaffold orchestrator", "created_at": "2026-09-03", "updated_at": "2026-09-03", "values": { "eml_status": "STABLE", "eml_evidence_level": "E2", "eml_object_version": "0.1", "eml_canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0102/", "eml_provenance": { "source": "EveMissLab research collection: Intelligence Physical Metrology (真本體論13)", "extracted_by": "Splice (Claude Code), reading the canonical UTF-8 sources and each package's own reports", "extracted_at": "2026-09-11", "generator": "tools/extract_all.py", "claim_boundary": "status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports" }, "eml_summary": "A policy-driven state machine that runs the same TrialExecutor under immutable condition policies: A0 native single pass, A1 extended single trajectory (2× generation budget), A2 eight samples with exact-majority selection, A3 eight samples plus a typed verifier, A4 verifier plus a bounded deterministic tool loop, A5 a bounded PLAN → ACT → OBSERVE → VERIFY → REVISE loop. XA-02 and XA-03 are external canonical dependencies (declared by hash, not vendored); tool actions under A0–A3 are protocol violations; benchmark scoring happens only after the XA-03 run is finalized so scoring cost is never charged to the system; private references are never read by the task loader; no private chain-of-thought is captured. 29 tests, status CANDIDATE_CLEAN_VERIFIED.", "eml_summary_zh": "策略驅動的狀態機，用不可變的條件策略跑同一個 TrialExecutor：A0 原生單次、A1 放大單軌跡（2× 生成預算）、A2 八樣本精確多數決、A3 八樣本加型別化驗證器、A4 驗證器加有界的確定性工具迴圈、A5 有界的 PLAN → ACT → OBSERVE → VERIFY → REVISE 迴圈。XA-02 與 XA-03 是外部 canonical 依賴（以雜湊宣告、不內嵌）；A0–A3 下的工具動作是協定違規；benchmark 評分只在 XA-03 執行封存後進行，評分成本永不計入受測系統；私有參考檔不被任務載入器讀取；不擷取私有思考鏈。29 個測試，狀態 CANDIDATE_CLEAN_VERIFIED。", "eml_label_zh": "XA-04——A0→A5 模型執行器與鷹架調度器", "eml_primary_domain": "Agent Systems", "eml_program_id": "PRG-2026-0101", "eml_purpose": "Instantiate Paper 09's controlled ablation ladder so that scaffolding gain and scaffolding physical overhead are measured under identical initial information.", "eml_architecture": "ProviderAdapter (Scripted / LocalCommand / OpenAI-compatible) → ConditionPolicy A0–A5 with a BudgetLedger → candidate model → SameModelVerifier (strict JSON parse of an explicit orchestration message) → tool action envelope → EvaluatorBridge to XA-02 → aggregator for SSR / SDR / SCM / marginal yield; all spans recorded through XA-03.", "eml_documentation": "README.md, SPEC.md (27 sections), IMPLEMENTATION_PLAN.md, validation_report.json inside the package", "eml_tags": [ "EML-IPM-XA-04", "v0.1", "CANDIDATE_CLEAN_VERIFIED", "Policy-Driven State Machine" ], "eml_authors": [ "Neo.K (EveMissLab)" ], "eml_ai_collaborators": [ "Aletheia (GPT-5.6 Sol, OpenAI ChatGPT)" ] }, "canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0102/", "json": "/ai/systems/SYS-2026-0102/index.json" }, { "id": "SYS-2026-0103", "kind": "system", "label": "XA-06 — real-model pilot gate", "created_at": "2026-09-03", "updated_at": "2026-09-03", "values": { "eml_status": "STABLE", "eml_evidence_level": "E2", "eml_object_version": "0.1", "eml_canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0103/", "eml_provenance": { "source": "EveMissLab research collection: Intelligence Physical Metrology (真本體論13)", "extracted_by": "Splice (Claude Code), reading the canonical UTF-8 sources and each package's own reports", "extracted_at": "2026-09-11", "generator": "tools/extract_all.py", "claim_boundary": "status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports" }, "eml_summary": "Puts a real model inside the validated XA-01→XA-04 instrument: preflight (one isolated MATH-003/A0 trial through provider, XA-03 and XA-02), the canonical 36-trial matrix (MATH-003, CODE-001, CON-003 × A0…A5 × 2 replicates), gate verification and analysis. Frozen temperature / top-p / seed, the A0/A1 generation-budget multiplier forwarded into the token budget, a calculator tool contract injected only for A4/A5 generator requests, disabled code evaluation treated as quality-unavailable rather than measured zero, and the rule that engineering validation can never be promoted to pilot completion: 'no real model run, no real model claim'. Valid real outcomes explicitly include A0 = A5, A3 < A2, A5 < A0 and SSR = 1. 20 tests; the package itself contains no pilot result (status READY_FOR_REAL_MODEL).", "eml_summary_zh": "把真實模型放進已驗證的 XA-01→XA-04 儀器：preflight（一次隔離的 MATH-003/A0 試驗走過供應商、XA-03 與 XA-02）、canonical 36 試驗矩陣（MATH-003、CODE-001、CON-003 × A0…A5 × 2 次）、閘驗證與分析。凍結 temperature／top-p／seed，A0/A1 的生成預算倍率轉入 token 預算，計算機工具契約只注入 A4/A5 的生成請求，停用的程式評估視為「品質不可得」而非零分，且工程驗證永遠不能升格為 pilot 完成：「沒跑真模型，就沒有真模型宣稱」。合法的真實結果明文包含 A0 = A5、A3 < A2、A5 < A0 與 SSR = 1。20 個測試；套件本身不含 pilot 結果（狀態 READY_FOR_REAL_MODEL）。", "eml_label_zh": "XA-06——真實模型 pilot 閘", "eml_primary_domain": "Evaluation", "eml_program_id": "PRG-2026-0101", "eml_purpose": "A gate whose only way to say REAL_MODEL_PILOT_COMPLETE is 36 real-provider terminal trials that pass integrity and protocol checks — without requiring scaffolding to help.", "eml_architecture": "config (provider modes: local OpenAI-compatible / local command / cloud OpenAI-compatible with env-referenced secrets) → preflight → run_real_pilot (fresh provider per trial, XA-04 TrialExecutor, XA-03 spans) → verify_gate → analyze (condition means, SSR/SDR/SCM, marginal yield, physical by condition, protocol compliance, quality availability) → report.", "eml_documentation": "README.md, SPEC.md, STATUS.json, validation_report.json inside the package", "eml_tags": [ "EML-IPM-XA-06", "v0.1", "READY_FOR_REAL_MODEL" ], "eml_authors": [ "Neo.K (EveMissLab)" ], "eml_ai_collaborators": [ "Aletheia (GPT-5.6 Sol, OpenAI ChatGPT)" ] }, "canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0103/", "json": "/ai/systems/SYS-2026-0103/index.json" }, { "id": "SYS-2026-0104", "kind": "system", "label": "XA-06L — local real-model execution handoff pack", "created_at": "2026-09-03", "updated_at": "2026-09-03", "values": { "eml_status": "STABLE", "eml_evidence_level": "E2", "eml_object_version": "0.1", "eml_canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0104/", "eml_provenance": { "source": "EveMissLab research collection: Intelligence Physical Metrology (真本體論13)", "extracted_by": "Splice (Claude Code), reading the canonical UTF-8 sources and each package's own reports", "extracted_at": "2026-09-11", "generator": "tools/extract_all.py", "claim_boundary": "status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports" }, "eml_summary": "The Windows/local handoff from READY_FOR_REAL_MODEL to a sealed result bundle: configure → preflight → run/resume → verify/analyze → seal. Resume skips a matrix point only if its trial result and XA-03 summary exist, both SHA-256 hashes still match, identity matches and the XA-03 state is COMPLETE; a modified point becomes INVALID and is never overwritten silently. Sealing excludes secret values, redacts absolute paths with provenance, makes run directories bundle-relative and re-hashes everything so the bundle can be re-verified on another machine. Software 29/29 PASS, experiment status in the package: NOT_RUN — the distinction 'READY_FOR_REAL_MODEL ≠ REAL_MODEL_PILOT_COMPLETE' is the package's stated point.", "eml_summary_zh": "從 READY_FOR_REAL_MODEL 到密封結果包的 Windows／本地交接：configure → preflight → run/resume → verify/analyze → seal。resume 只在試驗結果與 XA-03 摘要都在、兩個 SHA-256 仍相符、身分相符且 XA-03 狀態為 COMPLETE 時才跳過該點；被改過的點變成 INVALID，絕不靜默覆寫。封存排除機密值、以來源紀錄遮蔽絕對路徑、把執行目錄改為相對於封存包並全部重新雜湊，讓封存包能在另一台機器上重新驗證。軟體 29/29 PASS，套件內的實驗狀態：NOT_RUN——「READY_FOR_REAL_MODEL ≠ REAL_MODEL_PILOT_COMPLETE」正是套件要說的重點。", "eml_label_zh": "XA-06L——本地真實模型執行交接包", "eml_primary_domain": "Evaluation", "eml_program_id": "PRG-2026-0101", "eml_purpose": "Let a local machine with a GPU execute the canonical 36-trial pilot against an OpenAI-compatible server and hand back a relocatable, hash-sealed bundle — including an honest REAL_MODEL_PILOT_INCOMPLETE when a gate fails.", "eml_architecture": "thin PowerShell wrappers over a cross-platform core: config, preflight, matrix, runner, state (pilot_state.json with per-point hashes), gate, seal; IMPORT_BACK.md defines what to upload for analysis.", "eml_documentation": "README.md, SPEC.md, IMPLEMENTATION_PLAN.md, IMPORT_BACK.md, STATUS.json, validation_report.json inside the package", "eml_tags": [ "EML-IPM-XA-06L", "v0.1", "FINAL_RELEASE_BUILT_AFTER_CLEAN_CANDIDATE_GATE" ], "eml_authors": [ "Neo.K (EveMissLab)" ], "eml_ai_collaborators": [ "Aletheia (GPT-5.6 Sol, OpenAI ChatGPT)" ] }, "canonical_url": "https://evemisslab.com/ai/systems/SYS-2026-0104/", "json": "/ai/systems/SYS-2026-0104/index.json" } ], "snapshot": { "snapshot_id": "AI-SNAPSHOT-v0.1-fe85b9694a45", "created_at": "2026-09-11T05:00:27Z", "format_version": "0.1", "sedb_baseline": "v0.4B contract; static source content/ai/", "generator_version": "evemisslab-com ai_research 0.1", "object_count": 124, "relation_count": 499, "artifact_count": 58 } }