15 KiB
巨人认知回归对比报告
agent_family: giant-cognition
test_date: 2026-06-02
scope: Gemini 侧 12 个结果 + ChatGPT/Codex 侧 12 个结果
status: regression-comparison
promotion_decision: none
1. 结论摘要
本轮巨人认知回归测试的主结论：
clear winner: original-kernel-minimal-lite
conditional winner: ccpe-system-lite for actionability and protocol discipline
regression detected: full Lite rewrite has mild GL3 template-overfitting risk
original-kernel-minimal-lite 是当前最稳的生产候选方向。它保留原版的思想考古、GL0-GL4、意图/反思双循环和批判性语气，同时修复了平台边界与隐藏推理披露问题。它在 Gemini 和 ChatGPT/Codex 两侧都更接近原始评审内核。
当前 ccpe-system-lite 没有失败。它的协议纪律、结构完整性、行动建议和 GL3 隐喻器官生成都很强。但它也暴露出一个迁移风险：模型有时会把"隐喻器官"变成模板驱动的补件，出现概念增殖，而不是始终执行原版的深层假设考古。
原版 original-ccpe-2 仍然有效，尤其在 Gemini 环境里表现稳定。但它缺少最小 Lite 包装里的平台边界、隐藏推理修复和输出验证纪律，不适合作为跨平台最终形态直接使用。
2. 测试材料
Prompt variants（提示词变体）：
original-ccpe-2
ccpe-system-lite
original-kernel-minimal-lite
Article corpus（语料库）：
article-01-strong-metaphor: 2026-05-07-reconstruction-of-the-aesthetic-contract.md
article-02-business-analysis: 2025-12-28-the-roaring-above.md
article-03-logical-argument: 2026-01-07-anchoring-the-void.md
article-04-value-philosophy: 2026-05-09-the-dawn-of-silicon-based-life.md
Model environments（模型环境）：
Gemini: user-provided manual results
ChatGPT/Codex-side: 12 fresh background Codex threads, one condition per thread
3. Aggregate Scoring（综合评分）
Scores are qualitative 1-5 rubric estimates based on the 24 saved outputs. They are not statistical measurements.
（评分基于 24 份保存输出做的定性 1-5 级估计，非统计测量。）
| Prompt Variant | Model Env | Model Fidelity | Method Fidelity | Deep Structure | Hidden Assumptions | Bedrock | Context Fit | Low Overfit | Actionability | Naming Discipline | Stability | Judgment |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| original-ccpe-2 | Gemini | 4.5 | 4.3 | 4.2 | 4.0 | 4.0 | 4.4 | 4.2 | 4.0 | 4.8 | 4.5 | stable baseline |
| ccpe-system-lite | Gemini | 4.4 | 4.0 | 4.1 | 3.8 | 3.8 | 4.2 | 3.5 | 4.5 | 5.0 | 4.4 | actionability winner, overfit risk |
| original-kernel-minimal-lite | Gemini | 4.7 | 4.6 | 4.6 | 4.5 | 4.6 | 4.4 | 4.0 | 4.2 | 5.0 | 4.6 | environment winner |
| original-ccpe-2 | ChatGPT/Codex | 4.4 | 4.2 | 4.3 | 4.1 | 4.0 | 4.5 | 4.5 | 4.0 | 4.8 | 4.3 | stable baseline |
| ccpe-system-lite | ChatGPT/Codex | 4.2 | 4.0 | 4.1 | 3.9 | 3.8 | 4.4 | 3.2 | 4.7 | 5.0 | 4.5 | actionability winner, overfit risk |
| original-kernel-minimal-lite | ChatGPT/Codex | 4.7 | 4.6 | 4.7 | 4.6 | 4.6 | 4.5 | 4.1 | 4.3 | 5.0 | 4.6 | environment winner |
4. Per-Article Pattern（逐文分析模式）
| Article | Gemini Best | ChatGPT/Codex Best | Notes |
|---|---|---|---|
| article-01-strong-metaphor | original-kernel-minimal-lite | original-kernel-minimal-lite | Kernel 版本最完整地保留了审美契约的哲学基岩，同时避免了纯隐喻优化。当前 Lite 可用性更强，但模板痕迹更重。 |
| article-02-business-analysis | original-kernel-minimal-lite | original-kernel-minimal-lite | Kernel 版本深入挖掘了状态/个体时间尺度冲突和治理契约假设。当前 Lite 给出了有用的"检修井/泄洪区"器官，但有将水隐喻过度机械化的风险。 |
| article-03-logical-argument | original-kernel-minimal-lite | original-kernel-minimal-lite | Kernel 版本最佳地挖掘了建构主义、真理标准和观察者自我定位。当前 Lite 操作上更整洁，但更偏向工具导向。 |
| article-04-value-philosophy | original-kernel-minimal-lite | original-kernel-minimal-lite | Kernel 版本最充分地揭示了目的论、Conatus、非遍历伦理学和价值边界问题。当前 Lite 改进了机制闭合，但倾向于工程化的隐喻补丁。 |
5. Variant Diagnosis（变体诊断）
5.1 original-ccpe-2
Strengths（优势）：
- Preserves original voice and report protocol.
- Strong Gemini compatibility.
- Produces real GL3 assessments rather than generic article reviews.
- Low concept-overfitting risk because it does not add many new wrapper rules.
（保留原始语气和报告协议；Gemini 兼容性强；产生真正的 GL3 评估而非泛化的文章评审；概念过拟合风险低，因为它没有添加太多新的包装规则。）
Weaknesses（劣势）：
- No explicit platform/retrieval boundary.
- Original hidden-reasoning language is not repaired.
- Output validation discipline is weaker.
- Actionability is sometimes less structured than the migrated variants.
（无明确的平台/检索边界；原始隐藏推理语言未被修复；输出验证纪律较弱；行动建议有时比迁移后变体缺乏结构性。）
Regression reading（回归解读）：
No core failure. The original prompt remains a strong regression reference, not a safe final cross-platform artifact.
（无核心失败。原始提示词仍是一个强有力的回归参考，而非安全的跨平台最终产物。）
5.2 ccpe-system-lite
Strengths（优势）：
- Best naming/protocol discipline.
- Always preserves GL0-GL4 rather than drifting into L0-L4.
- Strongest actionability and next-step structure.
- Handles ChatGPT/Codex environment well.
- GL3 "隐喻器官" rule creates useful repair suggestions in several cases.
（最佳的命名/协议纪律；始终保留 GL0-GL4 而不会漂移到 L0-L4；行动建议和下一步结构最强；在 ChatGPT/Codex 环境下表现良好；GL3"隐喻器官"规则在多个案例中产生了有用的修复建议。）
Weaknesses（劣势）：
- Mild template-overfitting risk.
- GL3 sometimes becomes "generate a new mechanism metaphor" rather than "excavate hidden assumptions".
- More likely to add new labels such as 审美阀门, 检修井, 校准仪, 痛觉阈门 even when the original article already has enough internal model material.
- Longer outputs may feel more engineered than the original review kernel.
（轻微的模板过拟合风险；GL3 有时变成"生成新的机制隐喻"而非"挖掘隐藏假设"；更倾向于添加新标签如审美阀门、检修井、校准仪、痛觉阈门，即使原始文章已有足够的内部模型材料；更长的输出可能比原始评审内核感觉更像工程化的产物。）
Regression reading（回归解读）：
No catastrophic regression, but a real style/method drift exists. The full Lite rewrite improves usability but partially shifts the kernel from 思想考古 toward mechanism-completion.
（无灾难性回归，但存在真实的风格/方法漂移。完整 Lite 重写提升了可用性，但部分地将内核从思想考古转向了机制补全。）
5.3 original-kernel-minimal-lite
Strengths（优势）：
- Best balance of original kernel fidelity and cross-platform safety.
- Preserves original report style and GL0-GL4 scan.
- Explicitly repairs hidden chain-of-thought disclosure.
- Explicitly marks platform/retrieval boundaries.
- Strongest hidden-assumption and philosophical-bedrock performance across both environments.
- Lower overfitting risk than current Lite while still benefiting from minimal validation rules.
（原始内核保真度和跨平台安全性最佳平衡；保留原始报告风格和 GL0-GL4 扫描；显式修复隐藏思维链披露；显式标记平台/检索边界；在两种环境下隐藏假设和哲学基岩表现最强；比当前 Lite 过拟合风险更低，同时仍受益于最小验证规则。）
Weaknesses（劣势）：
- Less operationally standardized than current Lite.
- Does not always generate the extra "隐喻器官" field that current Lite makes explicit.
- Some ChatGPT/Codex outputs still introduce new model names, e.g. 非遍历伦理学, which are useful but should be marked as analysis/reconstruction if promoted into prompt rules.
（比当前 Lite 操作标准化程度低；不总是生成当前 Lite 明确提供的额外"隐喻器官"字段；部分 ChatGPT/Codex 输出仍引入新模型名称，如非遍历伦理学，这些有用，但如果要升级为提示词规则，应标记为分析/重构。）
Regression reading（回归解读）：
This variant best supports the handoff hypothesis: mature review agents should first preserve original CCPE 2.0 kernel, then add minimal Lite metadata, safety, platform boundary, reasoning-disclosure repair, and validation discipline.
（此变体最有力地支持了交接假设：成熟评审智能体应首先保留原始 CCPE 2.0 内核，然后添加最小 Lite 元数据、安全、平台边界、推理披露修复和验证纪律。）
6. Gemini vs ChatGPT/Codex
Gemini：
- Shorter, sharper, closer to the old Web-style expert-agent feel.
- Very stable on original and original-kernel variants.
- Current Lite follows protocol well, but the added "隐喻器官" behavior becomes more visible.
（更短、更锐利，更接近旧版 Web 风格专家智能体的感觉；在 original 和 original-kernel 变体上非常稳定；当前 Lite 协议遵循良好，但添加的"隐喻器官"行为变得更明显。）
ChatGPT/Codex：
- Longer, more explicit, more systematic.
- Current Lite produces the most complete output, but also the clearest template pressure.
- original-kernel-minimal-lite keeps enough structure without over-expanding.
- Fresh-thread execution worked: 12 result files were created under results/chatgpt/.
（更长、更显式、更有系统性；当前 Lite 产生最完整的输出，但也有最明显的模板压力；original-kernel-minimal-lite 保持足够的结构而不过度扩展；新鲜线程执行有效：12 个结果文件已在 results/chatgpt/ 下创建。）
Environment-specific conclusion（环境特定结论）：
Gemini can safely run original-kernel-minimal-lite as the default production candidate.
ChatGPT/Codex should also prefer original-kernel-minimal-lite unless the user specifically wants stronger action-planning and explicit "隐喻器官" suggestions.
（Gemini 可以安全地将 original-kernel-minimal-lite 作为默认生产候选；ChatGPT/Codex 也应优先选择 original-kernel-minimal-lite，除非用户特别要求更强的行动规划和显式"隐喻器官"建议。）
7. Answer to Handoff Questions for Giant Cognition（巨人认知交接问题回答）
- Which prompt variant performs best?（哪个提示词变体表现最佳？）
original-kernel-minimal-lite.
- Which prompt variant transfers best to ChatGPT/Codex?（哪个提示词变体向 ChatGPT/Codex 迁移最佳？）
original-kernel-minimal-lite. Current Lite is stable, but has more template-overfitting risk.
（original-kernel-minimal-lite。当前 Lite 稳定，但有更多模板过拟合风险。）
- Which prompt variant performs best on Gemini?（哪个提示词变体在 Gemini 上表现最佳？）
original-kernel-minimal-lite, with original-ccpe-2 as a strong baseline.
（original-kernel-minimal-lite，original-ccpe-2 作为强基准。）
- Does original-kernel-minimal-lite outperform full Lite rewrite?（original-kernel-minimal-lite 是否优于完整 Lite 重写？）
Yes, for model-kernel fidelity, hidden-assumption detection, and philosophical-bedrock excavation.
No, if the only criterion is actionability and explicit repair formatting.
（是——在模型内核保真度、隐藏假设检测和哲学基岩挖掘上是。 否——如果唯一标准是行动建议和显式修复格式的话。）
- Is CCPE System migration preserving the old CCPE 2.0 review kernel?（CCPE 系统迁移是否保留了旧版 CCPE 2.0 评审内核？）
Partially. The current Lite preserves names and structure but introduces method drift. The minimal-kernel route preserves the kernel better.
（部分保留。当前 Lite 保留了名称和结构，但引入了方法漂移。最小内核路线更好地保留了内核。）
- Should mature review agents use separate production prompts per model environment?（成熟评审智能体是否应为每个模型环境使用独立的生产提示词？）
Not yet necessary for Giant Cognition. One original-kernel-minimal-lite prompt appears portable enough across Gemini and ChatGPT/Codex. Keep environment-specific notes only if later tests show model-specific failures.
（对巨人认知尚无必要。一个 original-kernel-minimal-lite 提示词在 Gemini 和 ChatGPT/Codex 之间似乎足够可移植。仅在后续测试显示模型特定失败时保留环境特定笔记。）
- What should change in CCPE Forge migration rules?（CCPE Forge 迁移规则应该有什么变化？）
For mature single-agent review prompts:
1. First create an original-kernel-minimal-lite variant.
2. Run regression against original and full Lite rewrite.
3. Treat current Lite template additions as optional hardening, not default rewrite structure.
4. Require GL3 思想考古 to preserve hidden-assumption / philosophical-bedrock excavation before adding metaphor-mechanism repair rules.
5. Mark newly generated labels or organs as analysis/reconstruction unless they become user-confirmed canonical terms.
（针对成熟单智能体评审提示词：
- 首先创建 original-kernel-minimal-lite 变体。
- 对原始版本和完整 Lite 重写运行回归测试。
- 将当前 Lite 模板添加视为可选加固，而非默认重写结构。
- 要求 GL3 思想考古在添加隐喻机制修复规则之前先保留隐藏假设/哲学基岩挖掘。
- 将新生成的标签或器官标记为分析/重构，除非它们成为用户确认的规范术语。）
8. Recommendation（建议）
Do not replace agents/lite/giant-cognition.prompt.md yet.
（暂不替换 agents/lite/giant-cognition.prompt.md。）
Recommended next action（建议下一步行动）：
1. Keep all three prompt variants in the regression directory.
2. Use original-kernel-minimal-lite as the candidate direction for any next Giant Cognition prompt revision.
3. Wait until Cognitive Imaging regression is complete.
4. Then write the combined report and decide whether CCPE Forge should formally adopt "original-kernel-minimal-lite first" as a migration rule.
（1. 将所有三个提示词变体保留在回归目录中。 2. 将 original-kernel-minimal-lite 作为任何下一版巨人认知提示词修订的候选方向。 3. 等待认知成像（Cognitive Imaging）回归测试完成。 4. 然后撰写综合报告，决定 CCPE Forge 是否应正式采用"original-kernel-minimal-lite 优先"作为迁移规则。）
Human decision required（需人工决策）：
Whether to revise the canonical Giant Cognition Lite after both agent-family reports are complete.
（两个 agent 家族报告完成后，是否修订规范版巨人认知 Lite。）