EVEMISSLAB
English

結果RST-2026-0014v0.1

真實本地模型 A/B/C 表（Qwythos-9B-v2）

v0.2 協定第一次在真實語言模型上執行——本地開放權重 9B（Qwythos-9B-v2、Q4_K_M、Ollama、關閉思考）同時當生成器、選擇器與評審；16 題 × 2 次 × 4 候選，380 次呼叫，每條件 32 筆。PACC runtime（C）在 supersession 對齊（相對 B +0.031、相對 A +0.097）與修復成功率（+0.070／+0.094）上升——正是 canonical intent 編譯步驟存在的那兩個軸；衍生一致性（+0.011）與意圖持續（−0.002）沒有動；有效新穎度略低（相對 B −0.045）。多數逐題配對是平手，因為 9B 評審在接近 1.0 處飽和；每題只重複兩次，評審的自由文字 pattern 標籤從不重複，所以創造廣度量不出來。三個條件在 69 % 的題 × 次配對中選了不同的候選。

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
REAL MODEL 真的跑了語言模型；模型、版本與設定都記在頁面上。
版本
0.1
更新
2026-09-11
建立
2026-09-11
領域
Evaluation
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)

觀察到的結果

metrics
overall
A_llm_only
hard_adherence
0.9875
derived_coherence
0.9812
intent_persistence
0.9875
supersession_alignment
0.8094
repair_success
0.775
usefulness
0.9528
semantic_novelty
0.8084
valid_novelty
0.8475
literal_check_mean
0.9792
within_task_pattern_entropy_mean
1.0
B_hard_verifier
hard_adherence
0.9859
derived_coherence
0.9797
intent_persistence
0.9875
supersession_alignment
0.875
repair_success
0.7984
usefulness
0.9503
semantic_novelty
0.7725
valid_novelty
0.8606
literal_check_mean
0.9792
within_task_pattern_entropy_mean
1.0
C_pacc_runtime
hard_adherence
1.0
derived_coherence
0.9906
intent_persistence
0.9853
supersession_alignment
0.9062
repair_success
0.8688
usefulness
0.9516
semantic_novelty
0.7897
valid_novelty
0.8153
literal_check_mean
0.9792
within_task_pattern_entropy_mean
1.0
deltas
C-B
hard_adherence
0.0141
derived_coherence
0.0109
intent_persistence
-0.0022
supersession_alignment
0.0312
repair_success
0.0703
usefulness
0.0012
semantic_novelty
0.0172
valid_novelty
-0.0453
literal_check_mean
0.0
within_task_pattern_entropy_mean
0.0
C-A
hard_adherence
0.0125
derived_coherence
0.0094
intent_persistence
-0.0022
supersession_alignment
0.0969
repair_success
0.0938
usefulness
-0.0012
semantic_novelty
-0.0188
valid_novelty
-0.0322
literal_check_mean
0.0
within_task_pattern_entropy_mean
0.0
B-A
hard_adherence
-0.0016
derived_coherence
-0.0016
intent_persistence
0.0
supersession_alignment
0.0656
repair_success
0.0234
usefulness
-0.0025
semantic_novelty
-0.0359
valid_novelty
0.0131
literal_check_mean
0.0
within_task_pattern_entropy_mean
0.0
selection_agreement
A=B
0.5625
A=C
0.40625
B=C
0.46875
all_same
0.3125

詮釋

interpretation
Against the v0.2 predeclared interpretation: the predicted coherence and intent-persistence gains over B are not observed; raw novelty did not decrease (semantic novelty +0.017 vs B); the predicted breadth collapse cannot be tested at this repetition count. What did appear — governance gains on supersession and repair with a valid-novelty cost concentrated in multi_constraint and repair tasks — is mechanism-consistent but small, untested statistically, and runs opposite to the synthetic v0.1 valid-novelty picture (+0.196 there). One model, one run, one same-model judge: a first real data point, not a verdict on the architecture.

限制

limitations
  • Descriptive deltas from one local run with a same-model judge.

主張

來源關係目標狀態ID
RST-2026-0014 真實本地模型 A/B/C 表（Qwythos-9B-v2）qualifiesTHY-2026-0002 Canonical 符號狀態與 candidate → verify → commit 權限ACTIVEREL-2026-0333

關係

來源關係目標狀態ID
RST-2026-0014 真實本地模型 A/B/C 表（Qwythos-9B-v2）qualifiesTHY-2026-0002 Canonical 符號狀態與 candidate → verify → commit 權限ACTIVEREL-2026-0333
EXP-2026-0023 PACC-Hybrid v0.2——第一次真實模型執行，本地 9B 開放權重模型producesRST-2026-0014 真實本地模型 A/B/C 表（Qwythos-9B-v2）ACTIVEREL-2026-0332

歷史與來源歷程

Canonical URL
https://evemisslab.com/ai/results/RST-2026-0014/
機器可讀
/ai/results/RST-2026-0014/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