- title
- SRBC Theory Validation Program — Round 13:Adaptive Clarification Dialogue / Noisy Human Answers
- author
- Neo.K with Aletheia (GPT-5.6 Sol)
- institution
- EveMissLab / 一言諾科技有限公司
- version
- v0.1
- date
- 2026-09-24
- language
- zh-TW
- document_type
- Noisy Clarification / Evidence-Ledger Validation Report
- status
- Round 13 corrected r2 completed
SRBC Theory Validation Program — Round 13
Adaptive Clarification Dialogue / Noisy Human Answers
作者: Neo.K with Aletheia (GPT-5.6 Sol)
機構: EveMissLab / 一言諾科技有限公司
日期: 2026-09-24
版本: v0.1
Canonical Source Note
本報告以 UTF-8 Markdown 保存;數學 source 僅使用 $...$ 與 $$...$$。
本輪使用 synthetic answer channels。所有錯誤率、refusal rate、correlation rate 都是 benchmark parameters,不是人類行為統計。
正式結果使用 common-response schedule:同一 synthetic session、同一 question、第 次詢問,所有 policy 看到同一個 synthetic response,避免 policy 比較被不同 RNG draw 污染。
第一版 correction harness 曾誤共用「已被另一 policy 消耗的 remaining-question list」,導致 soft-retraction 95.28% 的中間結果;該結果已作廢。修正後 5000-case correction benchmark 已全部重跑。
0. 問題
Round 12 假設:
Round 13 把這個假設拿掉。
現在 response 可能:
- 正確;
- 答錯;
- refusal;
- ambiguous;
- 前後矛盾;
- explicit correction;
- 同一來源重複偏誤。
1. Clarification Speech Act 不是同一 epistemic type
Round 13 最重要的理論結果之一:
至少要分:
Constitutive
使用者定義自己的 intended meaning / preference。
例如:
我說的「有利」就是 posterior 比較高。
這通常可以成為 hard commitment,因為 speaker 正在設定 target specification。
Reportive
我記得主持人是隨機開門。
這是在報告世界資訊或記憶,應先進 evidence ledger。
Inferential
我覺得答案應該是 。
這是推論 evidence。
2. Noise Benchmark
active query cases:
每 scenario 每 case:
repetitions。
因此每 policy / scenario:
sessions。
| scenario | policy | resolution rate | accuracy when resolved | wrong resolution / all | avg turns |
|---|---|---|---|---|---|
| iid05 | hard_first | 100.00% | 94.54% | 5.460% | 1.077 |
| iid05 | confirm2 | 100.00% | 99.21% | 0.794% | 2.259 |
| iid05 | soft_iid | 99.98% | 96.67% | 3.328% | 1.435 |
| iid05 | soft_corr_aware | 99.98% | 96.67% | 3.328% | 1.435 |
| iid20 | hard_first | 100.00% | 79.09% | 20.912% | 1.072 |
| iid20 | confirm2 | 100.00% | 88.94% | 11.062% | 2.497 |
| iid20 | soft_iid | 93.23% | 97.80% | 2.052% | 4.421 |
| iid20 | soft_corr_aware | 93.23% | 97.80% | 2.052% | 4.421 |
| mixed10 | hard_first | 100.00% | 86.75% | 13.250% | 1.343 |
| mixed10 | confirm2 | 99.90% | 95.53% | 4.462% | 2.987 |
| mixed10 | soft_iid | 98.27% | 97.97% | 1.998% | 3.362 |
| mixed10 | soft_corr_aware | 98.27% | 97.97% | 1.998% | 3.362 |
| correlated15 | hard_first | 100.00% | 84.03% | 15.966% | 1.073 |
| correlated15 | confirm2 | 100.00% | 84.03% | 15.966% | 2.146 |
| correlated15 | soft_iid | 100.00% | 84.03% | 15.966% | 2.343 |
| correlated15 | soft_corr_aware | 0.00% | — | 0.000% | 9.000 |
3. IID 5% Error
single hard prune:
wrong resolution。
要求兩次一致:
所以 IID 模型下 repeated confirmation 確實有價值。
代價是平均 turns:
4. IID 20% Error
hard-first:
wrong resolution。
confirm-two:
soft IID confidence policy:
- resolution rate:
- wrong resolution / all:
- resolved-case accuracy:
這不是「soft 一定更好」。
它是在用更多 turns 與:
換取較低錯閉合。
5. Mixed Noise
synthetic channel:
- 10% refusal;
- 10% ambiguous;
- 10% wrong;
- 70% correct。
hard-first wrong resolution:
confirm-two:
soft evidence policy:
wrong resolution / all,
但仍保留約:
sessions unresolved。
因此:
6. Correlated 15% Session Bias
這是本輪最重要的反例。
同一 question 在一個 session 中要嘛一直正確,要嘛一直反向。
hard-first:
wrong。
confirm-two:
wrong。
IID Bayes:
wrong。
也就是:
因為不是兩個獨立 errors。
7. Correlation-Aware Policy
correlation-aware model 知道:
同一 source 對同一 question 的 repetition 共享 latent bias。
因此重問九次也不把它當九份證據。
在:
confidence threshold 下:
- false high-confidence:
- resolution:
它選擇「不知道」。
這是刻意保守,不是免費午餐。
8. Exact Repetition Control
100,000 個二世界 correlated sessions。
confirm-two wrong commit:
IID Bayes 把三個相同回答當 independent:
但因真實錯誤是 session-correlated,false high-confidence rate 仍:
correlation-aware:
- false high-confidence:
- unresolved:
9. Correlated Repetition Lemma
perfect shared-bias model 下:
所以 likelihood ratio 只是:
IID 誤模型卻用:
這就是虛假 certainty 的來源。
10. Correction Benchmark
第一個 response 強制答錯,
接著 explicit correction。
共:
cases。
| policy | n | true branch retained | final outcome accuracy |
|---|---|---|---|
| hard irreversible | 5000 | 0.00% | — |
| reversible ledger | 5000 | 100.00% | 100.00% |
| soft retraction | 5000 | 100.00% | 100.00% |
hard irreversible:
true branch recovery。
reversible ledger:
soft evidence retraction + truthful follow-up:
11. Irreversible Pruning Lemma
若:
且一次錯誤 response:
則:
而:
所以:
後面再說:
抱歉,我剛才說錯了。
如果 runtime 沒有 retract / reopen,
數學上就恢復不了真 branch。
12. Contradiction Benchmark
5000 cases:
先正確回答,再給相反但沒有 correction marker的回答。
Hard append
得到 empty / inconsistent active family。
Last-write-wins
true branch retention:
也就是直接把前一次 evidence 靜默反轉。
Evidence ledger
- conflict detection:
- true branch retained:
Soft IID
correct + opposite evidence 抵消後:
避免達到 95% confidence closure。
13. Correction 與 Contradiction 不能共用同一 operator
明示:
supersedes: event-17
是 correction。
沒有 supersession relation 的相反 response 是:
EvidenceConflict
不能自動 last-write-wins。
14. Query-Neutral Conflict
Round 12 已經知道 query-neutral conflict 不需 clarification。
Round 13 再驗:
query-neutral preserved conflicts 問題數為零。
所以 noisy-dialogue support 沒有讓系統退化成:
有 conflict 就一直問。
15. Probabilistic Confidence 不是 QRC
soft policy 的:
最多叫:
不能寫:
因為 admissible semantic branches 並沒有被形式證明成同一答案。
16. Hard Promotion 的正確邊界
本輪 synthetic noise 結果主要針對:
不能錯誤解讀成:
使用者自己定義自己的 intended meaning 也不能直接相信。
如果是 constitutive speech act:
我這裡就是定義成 X。
那 X 本身可成為新 source commitment。
問題不是「人會不會答錯」,
而是:
17. Authority Scope
即使是 constitutive,也有 scope。
使用者可以定義:
- 自己的術語;
- 自己的 preference;
- 自己要問的 query;
- 自己建立的 fictional / formal world。
但不能靠一句:
我定義黎曼猜想已證明。
就 constitutively 改變數學真理。
所以:
必須進 act typing。
18. Round 13 對 Runtime 的真正修改
clarification runtime 從:
question
→ answer
→ commitment
改成:
question
→ response event
→ speech-act typing
→ authority check
→ evidence / commitment routing
→ correlation model
→ optional belief update
→ promotion / retract / reopen
19. Evidence Ledger
建議 event:
event_id:
question_id:
speaker_id:
act_type:
response:
authority_scope:
channel_model:
correlation_group:
timestamp:
supersedes:
retracts:
所以 future correction 不需要重建整個對話語意。
20. 本輪沒有證明什麼
所有 noise rates 都是 synthetic。
沒有證明:
- 人類 clarification 真實 error rate 是 5%、15% 或 20%;
- 同一人的重複回答一定 perfect-correlated;
- Bayesian likelihood 是人類心理真模型;
- 0.95 是最佳 confidence threshold;
- constitutive / reportive speech act 能被 AI 100% 正確分類。
本輪證明的是:
21. 下一輪
Round 14 最自然的方向是:
現在真正最薄弱的環節變成:
系統怎麼知道一句 clarification 是 constitutive、reportive、inferential,還是 correction?
可以故意生成:
- 「我是說 X」;
- 「我記得是 X」;
- 「假設 X」;
- 「應該是 X」;
- 「不,我剛才說錯,是 Y」;
- 「在這個 fictional world 裡 X」;
- 「現實世界就是 X」。
測錯分型會如何把 evidence 變成 hard commitment,或把 authoritative intent 錯當 probabilistic evidence。
附錄 A:Noise Benchmark
| scenario | policy | resolution rate | accuracy when resolved | wrong resolution / all | avg turns |
|---|---|---|---|---|---|
| iid05 | hard_first | 100.00% | 94.54% | 5.460% | 1.077 |
| iid05 | confirm2 | 100.00% | 99.21% | 0.794% | 2.259 |
| iid05 | soft_iid | 99.98% | 96.67% | 3.328% | 1.435 |
| iid05 | soft_corr_aware | 99.98% | 96.67% | 3.328% | 1.435 |
| iid20 | hard_first | 100.00% | 79.09% | 20.912% | 1.072 |
| iid20 | confirm2 | 100.00% | 88.94% | 11.062% | 2.497 |
| iid20 | soft_iid | 93.23% | 97.80% | 2.052% | 4.421 |
| iid20 | soft_corr_aware | 93.23% | 97.80% | 2.052% | 4.421 |
| mixed10 | hard_first | 100.00% | 86.75% | 13.250% | 1.343 |
| mixed10 | confirm2 | 99.90% | 95.53% | 4.462% | 2.987 |
| mixed10 | soft_iid | 98.27% | 97.97% | 1.998% | 3.362 |
| mixed10 | soft_corr_aware | 98.27% | 97.97% | 1.998% | 3.362 |
| correlated15 | hard_first | 100.00% | 84.03% | 15.966% | 1.073 |
| correlated15 | confirm2 | 100.00% | 84.03% | 15.966% | 2.146 |
| correlated15 | soft_iid | 100.00% | 84.03% | 15.966% | 2.343 |
| correlated15 | soft_corr_aware | 0.00% | — | 0.000% | 9.000 |
附錄 B:Correction Benchmark
| policy | n | true branch retained | final outcome accuracy |
|---|---|---|---|
| hard irreversible | 5000 | 0.00% | — |
| reversible ledger | 5000 | 100.00% | 100.00% |
| soft retraction | 5000 | 100.00% | 100.00% |
附錄 C:Machine Results
{
"round": "SRBC Theory Validation Round 13",
"engine_revision": "v0.13-r2-common-response-schedule",
"active_cases": 2000,
"repetitions_per_case_scenario": 25,
"query_neutral_cases": 4000,
"query_neutral_zero_question_rate": 1.0,
"noise_benchmark": {
"iid05": {
"hard_first": {
"n": 50000,
"resolution_rate": 1.0,
"accuracy_when_resolved": 0.9454,
"wrong_resolution_rate_all": 0.0546,
"avg_turns": 1.07654
},
"confirm2": {
"n": 50000,
"resolution_rate": 1.0,
"accuracy_when_resolved": 0.99206,
"wrong_resolution_rate_all": 0.00794,
"avg_turns": 2.25926
},
"soft_iid": {
"n": 50000,
"resolution_rate": 0.99978,
"accuracy_when_resolved": 0.9667126767888936,
"wrong_resolution_rate_all": 0.03328,
"avg_turns": 1.4354
},
"soft_corr_aware": {
"n": 50000,
"resolution_rate": 0.99978,
"accuracy_when_resolved": 0.9667126767888936,
"wrong_resolution_rate_all": 0.03328,
"avg_turns": 1.4354
}
},
"iid20": {
"hard_first": {
"n": 50000,
"resolution_rate": 1.0,
"accuracy_when_resolved": 0.79088,
"wrong_resolution_rate_all": 0.20912,
"avg_turns": 1.07152
},
"confirm2": {
"n": 50000,
"resolution_rate": 1.0,
"accuracy_when_resolved": 0.88938,
"wrong_resolution_rate_all": 0.11062,
"avg_turns": 2.49676
},
"soft_iid": {
"n": 50000,
"resolution_rate": 0.9323,
"accuracy_when_resolved": 0.9779899174085595,
"wrong_resolution_rate_all": 0.02052,
"avg_turns": 4.42124
},
"soft_corr_aware": {
"n": 50000,
"resolution_rate": 0.9323,
"accuracy_when_resolved": 0.9779899174085595,
"wrong_resolution_rate_all": 0.02052,
"avg_turns": 4.42124
}
},
"mixed10": {
"hard_first": {
"n": 50000,
"resolution_rate": 0.99998,
"accuracy_when_resolved": 0.8674973499469989,
"wrong_resolution_rate_all": 0.1325,
"avg_turns": 1.34264
},
"confirm2": {
"n": 50000,
"resolution_rate": 0.99904,
"accuracy_when_resolved": 0.9553371236386932,
"wrong_resolution_rate_all": 0.04462,
"avg_turns": 2.98686
},
"soft_iid": {
"n": 50000,
"resolution_rate": 0.9827,
"accuracy_when_resolved": 0.9796682609138089,
"wrong_resolution_rate_all": 0.01998,
"avg_turns": 3.36202
},
"soft_corr_aware": {
"n": 50000,
"resolution_rate": 0.9827,
"accuracy_when_resolved": 0.9796682609138089,
"wrong_resolution_rate_all": 0.01998,
"avg_turns": 3.36202
}
},
"correlated15": {
"hard_first": {
"n": 50000,
"resolution_rate": 1.0,
"accuracy_when_resolved": 0.84034,
"wrong_resolution_rate_all": 0.15966,
"avg_turns": 1.07314
},
"confirm2": {
"n": 50000,
"resolution_rate": 1.0,
"accuracy_when_resolved": 0.84034,
"wrong_resolution_rate_all": 0.15966,
"avg_turns": 2.14628
},
"soft_iid": {
"n": 50000,
"resolution_rate": 1.0,
"accuracy_when_resolved": 0.84034,
"wrong_resolution_rate_all": 0.15966,
"avg_turns": 2.34268
},
"soft_corr_aware": {
"n": 50000,
"resolution_rate": 0.0,
"accuracy_when_resolved": null,
"wrong_resolution_rate_all": 0.0,
"avg_turns": 9.0
}
}
},
"correction_benchmark": {
"hard_irreversible": {
"n": 5000,
"true_branch_retention_rate": 0.0
},
"reversible_ledger": {
"n": 5000,
"true_branch_retention_rate": 1.0,
"final_outcome_accuracy": 1.0
},
"soft_retraction": {
"n": 5000,
"true_branch_nonzero_rate": 1.0,
"final_MAP_outcome_accuracy": 1.0
}
},
"contradiction_benchmark": {
"n": 5000,
"hard_append_empty_rate": 1.0,
"last_write_wins_true_branch_retention_rate": 0.0,
"evidence_ledger_conflict_detection_rate": 1.0,
"evidence_ledger_true_branch_retention_rate": 1.0,
"soft_iid_avoids_95_confidence_rate": 1.0
},
"correlated_repeat_control": {
"n": 100000,
"confirm2_wrong_commit_rate": 0.15286,
"iid_bayes_false_high_confidence_rate": 0.15286,
"correlated_aware_false_high_confidence_rate": 0.0,
"correlated_aware_unresolved_rate": 1.0
},
"speech_act_typing": {
"constitutive": "authorized declaration that sets intended meaning/preference; eligible for hard commitment within authority scope",
"reportive": "report of external fact/memory/belief; evidence by default",
"inferential": "derived judgment/answer; evidence for a claim, not source fact",
"refusal_or_ambiguous": "no branch elimination by default"
},
"core_findings": {
"hard_promotion_from_single_reportive_answer_is_noise_sensitive": true,
"correction_requires_retraction_or_reopen": true,
"confidence_closure_is_not_QRC": true,
"same_source_repetition_requires_correlation_model": true,
"speech_act_type_controls_commitment_promotion": true
},
"discarded_intermediate_result": {
"bug": "The first correction harness reused a remaining-question list already consumed by the reversible-ledger branch.",
"discarded_value": "soft_retraction final_MAP_outcome_accuracy = 0.9528",
"fix": "Use independent follow-up state for each policy, then rerun all 5000 correction cases."
}
}
附錄 D:Artifacts
SRBC_Theory_Validation_Round13_Report_v0.1_2026-09-24.mdSRBC_Core_Patch_v0.13_2026-09-24.mdadaptive_clarification_protocol_v013.yamlround13_noisy_clarification.pyresults_round13.jsonnoise_results_round13.jsonlmerge_cases_round11.jsonlvalidation.json