資料集DAT-2026-0002v0.1
PACC-Hybrid v0.1 共享候選池
1,568 個候選池實例(7 類 × 28 題 × 8 次重複,每池 112 個候選),合成生成並一模一樣地提供給 A/B/C 三個條件,附 primary、multiseed 與 elastic 診斷的結果 JSON。不含任何語言模型輸出。
內容
description- pacc_hybrid_v0.1_primary.json, pacc_hybrid_v0.1_multiseed.json, pacc_hybrid_v0.1_elastic_multiseed.json, pacc_hybrid_v0.1_elastic_diagnostic.json inside the v0.1 bundle.
如何建立
generation_method- Seeded synthetic candidate generation (seed 20260909 primary; 8 fixed secondary seeds).
已知偏誤
known_bias- Synthetic candidates and synthetic scoring; a real-model pool is the v0.2 open item.
格式與規模
format- JSON
關係
| 來源 | 關係 | 目標 | 狀態 | ID |
|---|---|---|---|---|
DAT-2026-0002 PACC-Hybrid v0.1 共享候選池 | supports | RES-2026-0004 PACC 式 runtime 用在語言模型上:推理、意圖與創造廣度 | ACTIVE | REL-2026-0093 |
EXP-2026-0021 PACC-Hybrid v0.1——合成 A/B/C 架構見證 | uses_dataset | DAT-2026-0002 PACC-Hybrid v0.1 共享候選池 | ACTIVE | REL-2026-0315 |
歷史與來源歷程
- Canonical URL
- https://evemisslab.com/ai/data/DAT-2026-0002/
- 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