實驗EXP-2026-0012v0.1
PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？
豐富的關係圖比 v0.4 的精簡 bundle 保留多得多的來源／成對／模式資訊，且任務行動完全相同。它救不回凍結的 direct affine-sigmoid 到共同因座標的映射：精簡 0、豐富 0 次穩健通過，真實／控制比 ≈ 0.96–1.02，四個次要 seed 也沒有任何系統被救回。反事實來源翻轉探針常單獨通過，顯示局部方向對齊部分存在，但沒有全域可判別的座標。
假設
hypothesis- The v0.4 failure was information loss in the compact state; a richer non-probabilistic relation state admits the direct low-complexity latent coordinate.
設定
model_idsdataset_idsbenchmark_idssoftware_environment- Python; deterministic seeded generators; no network, no LLM.
執行
run_count- 2
random_seeds- 20260909
- 4 secondary seeds
controls- shuffled-target mapping
- constant prediction
- broken composition (v0.6+)
- target-local refit (v0.7+)
metricsverdict- RICH_STATE_DOES_NOT_RESCUE_LATENT_DEPENDENCE_COORDINATE
compact_robust_pass- 0
rich_robust_pass- 0
task_action_invariance- true
moderate_N0compact_dependence_js- 0.0605
rich_dependence_js- 0.06146
real_over_control- 0.999
詮釋
interpretation- Richer raw relation state is insufficient for the tested direct low-complexity latent coordinate — not that the coordinate is information-theoretically unrecoverable. Points to a coordinate-composition problem (v0.6).
限制
limitations- A direct affine-sigmoid observer may be too restrictive even when the raw state is informative.
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-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | runs_on | SYS-2026-0002 PACC-Lab——收斂猜想的微型實驗室 | ACTIVE | REL-2026-0206 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | uses_benchmark | BEN-2026-0001 PACC 微型實驗室協定 v0.1（凍結門檻） | ACTIVE | REL-2026-0207 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | uses_dataset | DAT-2026-0001 PACC 合成證據世界 | ACTIVE | REL-2026-0208 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | uses_model | MOD-2026-0001 N0——帶號支持 | ACTIVE | REL-2026-0209 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | uses_model | MOD-2026-0002 N1——序數錦標賽 | ACTIVE | REL-2026-0210 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | uses_model | MOD-2026-0003 N2——帶號圖 | ACTIVE | REL-2026-0211 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | uses_model | MOD-2026-0004 N3——約束競爭 | ACTIVE | REL-2026-0212 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | uses_model | MOD-2026-0005 Bayesian 參考（精確後驗／Beta-Bernoulli／聯合共同因／HMM） | ACTIVE | REL-2026-0213 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | tests | THY-2026-0005 PACC 猜想——四層收斂階梯 | ACTIVE | REL-2026-0214 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | extends | EXP-2026-0011 PACC-Lab v0.4——相關來源與依賴幾何 | ACTIVE | REL-2026-0215 |
EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | produced | ART-2026-0025 PACC-Lab v0.5 Representation Sufficiency FINAL artifact://evemisslab/adaptive-epistemic-systems/PACC-Lab_v0.5_Representation_Sufficiency_FINAL.zip | ACTIVE | REL-2026-0216 |
EXP-2026-0013 PACC-Lab v0.6——階層式組合座標 | extends | EXP-2026-0012 PACC-Lab v0.5——更豐富的關係狀態能救回潛在座標嗎？ | ACTIVE | REL-2026-0226 |
歷史與來源歷程
- Canonical URL
- https://evemisslab.com/ai/experiments/EXP-2026-0012/
- 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