recursive-decision-ledger
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseRecursive Decision Ledger
Recursive Decision Ledger
Use this skill when the user is trying to force deeper computation through
repeated rollouts or "Prime Gauss" style recursive prompting. Preserve the useful
part: repeated trials, prior memory, fresh information, and explicit marks.
Remove the unsafe part: pretending the loop proves certainty.
当用户试图通过重复推演或“Prime Gauss”式递归提示来实现更深层次的计算时,可使用此技能。保留有用的部分:重复试验、历史记忆、新信息和明确标记。移除不安全的部分:假设循环能证明确定性。
Ledger Contract
账本约定
Every rollout should record:
- rollout id and timestamp;
- prior accepted winner and prior watchlist;
- fresh information ingested;
- search space size;
- model families or heuristics used;
- trial count and effective trial count;
- top candidates;
- decision marks;
- coherence marks against the prior ledger;
- promotion gate result.
Prefer JSONL for append-only ledgers and Markdown for human summaries.
每次推演都应记录:
- 推演ID和时间戳;
- 之前认可的最优结果和关注列表;
- 纳入的新信息;
- 搜索空间规模;
- 使用的模型族或启发式方法;
- 试验次数和有效试验次数;
- 候选最优结果;
- 决策标记;
- 与历史账本的一致性标记;
- 晋升审核结果。
仅追加式账本优先使用JSONL格式,面向人类的摘要优先使用Markdown格式。
Rollout Loop
推演循环
- Load the prior ledger.
- Capture new information at time-step zero.
- Run the bounded search.
- Mark each candidate: accept, watch, reject, decay watch, or needs replay.
- Compare winners against prior winners and latest marked rollout.
- Downgrade candidates when drift, tail risk, stale data, or failed replay invalidates the previous mark.
- Append artifacts before summarizing.
- 加载历史账本。
- 在时间步零捕获新信息。
- 执行有界搜索。
- 为每个候选结果标记:认可、关注、拒绝、降低关注等级或需重放验证。
- 将当前最优结果与历史最优结果及最新标记的推演结果进行对比。
- 当出现数据漂移、尾部风险、数据过期或重放验证失败导致之前的标记失效时,降低候选结果的等级。
- 在生成摘要前追加相关产物。
Coherence Mark
一致性标记
Include a compact coherence mark:
text
Ensemble matches prior winner: true
Recursive matches prior winner: false
Latest rollout match: true
Live promotion allowed: false
Reason: replay and freshness gates not satisfied需包含简洁的一致性标记:
text
Ensemble matches prior winner: true
Recursive matches prior winner: false
Latest rollout match: true
Live promotion allowed: false
Reason: replay and freshness gates not satisfiedPromotion Rules
晋升规则
For trading, capital allocation, production deploys, migrations, or destructive
ops, recursive confidence is not approval.
Default to paper, dry-run, read-only, preview, or staged mode unless the user
explicitly approves the live action and the repo/service gate supports it.
Promote only when:
- the candidate beats the prior accepted winner on the chosen metric;
- correctness and replay checks pass;
- risk limits are explicit;
- the evidence is durable;
- the user has approved the live step when needed.
对于交易、资本配置、生产部署、迁移或破坏性操作,递归得出的置信度不等同于批准。
默认采用模拟交易、试运行、只读、预览或分阶段模式,除非用户明确批准执行实时操作且代码库/服务审核机制支持。
仅在满足以下条件时才可执行晋升:
- 候选结果在选定指标上优于之前认可的最优结果;
- 通过正确性和重放验证检查;
- 明确风险限制;
- 证据具备持久性;
- 必要时用户已批准实时步骤。
Summary Shape
摘要格式
Lead with the decision, not the drama:
text
Rollout 15 complete. The prior winner still holds, but edge deteriorated 17%.
Status: watch, not live. Next gate: 20 replay fills with fresh orderbook age
below threshold.摘要应直接呈现决策,而非过程细节:
text
Rollout 15 complete. The prior winner still holds, but edge deteriorated 17%.
Status: watch, not live. Next gate: 20 replay fills with fresh orderbook age below threshold.