實驗EXP-2026-0008v0.1

PACC-Lab v0.1——E0–E4：第一個微型見證

三個 primitive 層的非概率系統（N0 帶號支持、N1 序數錦標賽、N2 帶號圖）在四個隱藏假設上整合證據，旁邊是精確的 Bayesian 參考。三者在均勻與異質證據品質下都通過 Level-3 微型準則——held-out JS ≤ 0.0101、更新 JS ≤ 0.0042、干預 JS ≤ 0.0058、行動一致度 ≥ 0.957——shuffled-target 控制組約 0.31–0.33。冗餘診斷把 N0 與 N1 合併為一個精確重參數化家族，獨立家族只剩兩個，未達 PACC-A 最低要求三個。

研究狀態: STABLE 目前的研究結論相對穩定
證據等級: E3 重複實驗
結果: MIXED
資料基礎: SYNTHETIC 合成數據與理論推理。現在很多人把合成數據當成真的；這個實驗室刻意反過來說——在真正的混合模型出現之前，推論就只是推論，理論上可能不等於實際上可能。
版本: 0.1
更新: 2026-09-08
建立: 2026-09-08
領域: Model Representation, Evaluation, Formal AI
計畫: PRG-2026-0001 自適應世界狀態系統的第一原理框架
作者: Neo.K (EveMissLab)
AI 協作: Sol (GPT-5.6, OpenAI ChatGPT)

假設

hypothesis: A system whose canonical state and update rules do not require probability can converge toward a Bayesian reference in behaviour, held-out state representation and update dynamics.

設定

model_ids: MOD-2026-0001 N0——帶號支持
MOD-2026-0002 N1——序數錦標賽
MOD-2026-0003 N2——帶號圖
MOD-2026-0004 N3——約束競爭
MOD-2026-0005 Bayesian 參考（精確後驗／Beta-Bernoulli／聯合共同因／HMM）
dataset_ids: DAT-2026-0001 PACC 合成證據世界
benchmark_ids: BEN-2026-0001 PACC 微型實驗室協定 v0.1（凍結門檻）
software_environment: Python; deterministic seeded generators; no network, no LLM.

執行

run_count

random_seeds

20260908
8 fixed secondary seeds (post-hoc)

controls

shuffled-target mapping
constant prediction
broken composition (v0.6+)
target-local refit (v0.7+)

metrics

verdict

PRELIMINARY_PACC_B_R_D_MICRO_WITNESS_WITHOUT_STRONG_ATTRACTOR_CLOSURE

e0_purity

true

uniform

N0

agreement: 0.983073
heldout_js: 0.000192
update_js: 5.2e-05
intervention_js: 6.1e-05
shuffled_js: 0.322816

N1

agreement: 0.983073
heldout_js: 0.000192
update_js: 5.2e-05
intervention_js: 6.8e-05
shuffled_js: 0.312444

N2

agreement: 0.957465
heldout_js: 0.008053
update_js: 0.003872
intervention_js: 0.005756
shuffled_js: 0.309444

heterogeneous

N0

agreement: 0.983073
heldout_js: 0.002878
update_js: 0.000559
intervention_js: 0.000645

N2

agreement: 0.958767
heldout_js: 0.010109
update_js: 0.004186
intervention_js: 0.004187

independent_families

strong_attractor_minimum

secondary_8_seed_pass_rate

1.0

詮釋

interpretation: Non-probabilistic primitives can exhibit probability-like state and update structure in controlled micro-tasks; two independent convergent families are not enough to establish a computational attractor. Under uniform reliability additive signed support is closely related to rescaled log-evidence accumulation — the heterogeneous stress makes the result less trivial, not deeply equivalent.

限制

limitations: Does not show probability is false, Bayesian inference unnecessary, LLM internals equivalent, non-probabilistic systems superior, or probability an observer projection.
E4 dynamic-world numbers are descriptive only (HMM hazard and decay untuned).

重現

reproduction_instructions: Extract the version's FINAL bundle; python -m pytest -q; run the version's primary script with the recorded seed; docs/PACC_LAB_v0.N_RESULTS.md and docs/EXPERIMENT_PROTOCOL_v0.N.md are inside the bundle.

結果

來源	關係	目標	狀態	ID
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`produces`	`RST-2026-0006` v0.1 主要表：N0/N1/N2 皆 Level 3，兩個家族	`ACTIVE`	`REL-2026-0166`

記錄欄位

completed_at: 2026-09-08

關係

來源	關係	目標	狀態	ID
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`runs_on`	`SYS-2026-0002` PACC-Lab——收斂猜想的微型實驗室	`ACTIVE`	`REL-2026-0156`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`uses_benchmark`	`BEN-2026-0001` PACC 微型實驗室協定 v0.1（凍結門檻）	`ACTIVE`	`REL-2026-0157`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`uses_dataset`	`DAT-2026-0001` PACC 合成證據世界	`ACTIVE`	`REL-2026-0158`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`uses_model`	`MOD-2026-0001` N0——帶號支持	`ACTIVE`	`REL-2026-0159`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`uses_model`	`MOD-2026-0002` N1——序數錦標賽	`ACTIVE`	`REL-2026-0160`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`uses_model`	`MOD-2026-0003` N2——帶號圖	`ACTIVE`	`REL-2026-0161`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`uses_model`	`MOD-2026-0004` N3——約束競爭	`ACTIVE`	`REL-2026-0162`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`uses_model`	`MOD-2026-0005` Bayesian 參考（精確後驗／Beta-Bernoulli／聯合共同因／HMM）	`ACTIVE`	`REL-2026-0163`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`tests`	`THY-2026-0005` PACC 猜想——四層收斂階梯	`ACTIVE`	`REL-2026-0164`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`produced`	`ART-2026-0021` PACC-Lab v0.1 E0-E4 FINAL `artifact://evemisslab/adaptive-epistemic-systems/PACC-Lab_v0.1_E0-E4_FINAL.zip`	`ACTIVE`	`REL-2026-0165`
`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`produces`	`RST-2026-0006` v0.1 主要表：N0/N1/N2 皆 Level 3，兩個家族	`ACTIVE`	`REL-2026-0166`
`EXP-2026-0009` PACC-Lab v0.2——第三個家族（N3）與對抗性證據幾何	`extends`	`EXP-2026-0008` PACC-Lab v0.1——E0–E4：第一個微型見證	`ACTIVE`	`REL-2026-0178`

歷史與來源歷程

Canonical URL

https://evemisslab.com/ai/experiments/EXP-2026-0008/

機器可讀

/ai/experiments/EXP-2026-0008/index.json

快照

AI-SNAPSHOT-v0.1-fe85b9694a45

來源歷程

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