實驗EXP-2026-0019v0.1

PACC-Lab v0.12——有界二次轉換

轉換容量以一個預登記的步驟從 20 個仿射係數升到 60 個固定二次係數，chart、支撐、切分、控制與門檻不變。沒有家族被救回：反向保真度 0.0111（N0/N1）、0.0139（N2）、0.0141（N3），門檻 0.01，次要 0/4、較大規模 0/3；正向 cocycle 與反向一致性在二次類下存活。實驗室在此停止，而不是看到結果後升級到三次、四次或神經轉換。

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: NEGATIVE
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: 2026-09-09
建立: 2026-09-09
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: The reverse destination bias is an artefact of the 4-D affine transition class.

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

20260909
3
5
7
11
3 primary-scale-ish seeds

controls

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

metrics

verdict

REVERSE BIAS PERSISTS UNDER BOUNDED QUADRATIC

quadratic_reverse_fidelity

N0: 0.01108
N2: 0.01392
N3: 0.0141

gate

0.01

quadratic_reverse_pass

0/4 and 0/3

forward_cocycle

PASS

inverse_pairs

PASS (N2 2/3 at scale)

詮釋

interpretation: Rejects the affine-limitation explanation and strengthens a persistent directional/base-point mismatch as the next hypothesis. A sufficiently flexible approximator could fit any finite sample, which is exactly why capacity was bounded in advance.

限制

limitations: Does not prove that no nonlinear transition can remove the bias.

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.

記錄欄位

completed_at: 2026-09-09

關係

來源	關係	目標	狀態	ID
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`runs_on`	`SYS-2026-0002` PACC-Lab——收斂猜想的微型實驗室	`ACTIVE`	`REL-2026-0291`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`uses_benchmark`	`BEN-2026-0001` PACC 微型實驗室協定 v0.1（凍結門檻）	`ACTIVE`	`REL-2026-0292`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`uses_dataset`	`DAT-2026-0001` PACC 合成證據世界	`ACTIVE`	`REL-2026-0293`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`uses_model`	`MOD-2026-0001` N0——帶號支持	`ACTIVE`	`REL-2026-0294`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`uses_model`	`MOD-2026-0002` N1——序數錦標賽	`ACTIVE`	`REL-2026-0295`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`uses_model`	`MOD-2026-0003` N2——帶號圖	`ACTIVE`	`REL-2026-0296`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`uses_model`	`MOD-2026-0004` N3——約束競爭	`ACTIVE`	`REL-2026-0297`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`uses_model`	`MOD-2026-0005` Bayesian 參考（精確後驗／Beta-Bernoulli／聯合共同因／HMM）	`ACTIVE`	`REL-2026-0298`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`tests`	`THY-2026-0005` PACC 猜想——四層收斂階梯	`ACTIVE`	`REL-2026-0299`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`extends`	`EXP-2026-0018` PACC-Lab v0.11——雙向 cocycle 與反向一致性	`ACTIVE`	`REL-2026-0300`
`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`produced`	`ART-2026-0032` PACC-Lab v0.12 Bounded Nonlinear Transition FINAL `artifact://evemisslab/adaptive-epistemic-systems/PACC-Lab_v0.12_Bounded_Nonlinear_Transition_FINAL.zip`	`ACTIVE`	`REL-2026-0301`
`EXP-2026-0020` PACC-Lab v0.13——反向殘差場／基點依賴	`extends`	`EXP-2026-0019` PACC-Lab v0.12——有界二次轉換	`ACTIVE`	`REL-2026-0311`

歷史與來源歷程

Canonical URL

https://evemisslab.com/ai/experiments/EXP-2026-0019/

機器可讀

/ai/experiments/EXP-2026-0019/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