math-skill 2.1.0 → 3.0.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.en-US.md +159 -200
- package/README.md +159 -200
- package/agents/math-critic.en.md +2 -2
- package/agents/math-critic.md +8 -8
- package/commands/ask.md +3 -10
- package/design-patterns/attention/equivariant-attention.en.md +72 -0
- package/design-patterns/attention/equivariant-attention.md +72 -0
- package/design-patterns/attention/geometry-aware-attention.en.md +73 -0
- package/design-patterns/attention/geometry-aware-attention.md +72 -0
- package/design-patterns/attention/information-bottleneck-attention.en.md +72 -0
- package/design-patterns/attention/information-bottleneck-attention.md +72 -0
- package/design-patterns/attention/projection-attention.en.md +68 -0
- package/design-patterns/attention/projection-attention.md +68 -0
- package/design-patterns/attention/spectral-attention.en.md +71 -0
- package/design-patterns/attention/spectral-attention.md +71 -0
- package/design-patterns/compression/leverage-score-selection.en.md +71 -0
- package/design-patterns/compression/leverage-score-selection.md +71 -0
- package/design-patterns/compression/low-rank-kv-cache.en.md +60 -0
- package/design-patterns/compression/low-rank-kv-cache.md +60 -0
- package/design-patterns/compression/spectral-token-pruning.en.md +61 -0
- package/design-patterns/compression/spectral-token-pruning.md +61 -0
- package/design-patterns/compression/topology-preserving-compression.en.md +64 -0
- package/design-patterns/compression/topology-preserving-compression.md +64 -0
- package/design-patterns/loss/constraint-penalty.en.md +67 -0
- package/design-patterns/loss/constraint-penalty.md +78 -0
- package/design-patterns/loss/contrastive-loss.en.md +59 -0
- package/design-patterns/loss/contrastive-loss.md +68 -0
- package/design-patterns/loss/information-bottleneck-loss.en.md +69 -0
- package/design-patterns/loss/information-bottleneck-loss.md +74 -0
- package/design-patterns/loss/orthogonality-loss.en.md +56 -0
- package/design-patterns/loss/orthogonality-loss.md +60 -0
- package/design-patterns/loss/variational-loss.en.md +61 -0
- package/design-patterns/loss/variational-loss.md +71 -0
- package/design-patterns/representation/equivariant-split.en.md +87 -0
- package/design-patterns/representation/equivariant-split.md +89 -0
- package/design-patterns/representation/manifold-representation.en.md +85 -0
- package/design-patterns/representation/manifold-representation.md +87 -0
- package/design-patterns/representation/shared-private-decomposition.en.md +84 -0
- package/design-patterns/representation/shared-private-decomposition.md +86 -0
- package/design-patterns/representation/subspace-alignment.en.md +90 -0
- package/design-patterns/representation/subspace-alignment.md +92 -0
- package/design-patterns/routing/graph-routing.en.md +79 -0
- package/design-patterns/routing/graph-routing.md +81 -0
- package/design-patterns/routing/moe-routing.en.md +85 -0
- package/design-patterns/routing/moe-routing.md +87 -0
- package/design-patterns/routing/optimal-transport-routing.en.md +68 -0
- package/design-patterns/routing/optimal-transport-routing.md +72 -0
- package/design-patterns/routing/spectral-clustering-routing.en.md +88 -0
- package/design-patterns/routing/spectral-clustering-routing.md +90 -0
- package/knowledge-base/differential-geometry/connection.en.md +51 -0
- package/knowledge-base/differential-geometry/connection.md +51 -0
- package/knowledge-base/differential-geometry/curvature.en.md +49 -0
- package/knowledge-base/differential-geometry/curvature.md +49 -0
- package/knowledge-base/differential-geometry/geodesic.en.md +51 -0
- package/knowledge-base/differential-geometry/geodesic.md +51 -0
- package/knowledge-base/differential-geometry/manifold.en.md +47 -0
- package/knowledge-base/differential-geometry/manifold.md +47 -0
- package/knowledge-base/differential-geometry/metric-tensor.en.md +50 -0
- package/knowledge-base/differential-geometry/metric-tensor.md +50 -0
- package/knowledge-base/differential-geometry/tangent-space.en.md +48 -0
- package/knowledge-base/differential-geometry/tangent-space.md +48 -0
- package/knowledge-base/information-geometry/fisher-metric.en.md +56 -0
- package/knowledge-base/information-geometry/fisher-metric.md +56 -0
- package/knowledge-base/information-geometry/natural-gradient.en.md +54 -0
- package/knowledge-base/information-geometry/natural-gradient.md +54 -0
- package/knowledge-base/lie-theory/equivariance.en.md +54 -0
- package/knowledge-base/lie-theory/equivariance.md +54 -0
- package/knowledge-base/lie-theory/group-action.en.md +51 -0
- package/knowledge-base/lie-theory/group-action.md +51 -0
- package/knowledge-base/lie-theory/lie-algebra.en.md +51 -0
- package/knowledge-base/lie-theory/lie-algebra.md +51 -0
- package/knowledge-base/lie-theory/lie-group.en.md +51 -0
- package/knowledge-base/lie-theory/lie-group.md +51 -0
- package/knowledge-base/lie-theory/representation.en.md +52 -0
- package/knowledge-base/lie-theory/representation.md +52 -0
- package/knowledge-base/matrix-analysis/low-rank-approximation.en.md +49 -0
- package/knowledge-base/matrix-analysis/low-rank-approximation.md +49 -0
- package/knowledge-base/matrix-analysis/matrix-perturbation.en.md +49 -0
- package/knowledge-base/matrix-analysis/matrix-perturbation.md +49 -0
- package/knowledge-base/matrix-analysis/positive-semidefinite.en.md +50 -0
- package/knowledge-base/matrix-analysis/positive-semidefinite.md +50 -0
- package/knowledge-base/matrix-analysis/projection.en.md +47 -0
- package/knowledge-base/matrix-analysis/projection.md +47 -0
- package/knowledge-base/matrix-analysis/spectral-decomposition.en.md +48 -0
- package/knowledge-base/matrix-analysis/spectral-decomposition.md +48 -0
- package/knowledge-base/optimization/constrained-optimization.en.md +53 -0
- package/knowledge-base/optimization/constrained-optimization.md +53 -0
- package/knowledge-base/optimization/convex-optimization.en.md +50 -0
- package/knowledge-base/optimization/convex-optimization.md +50 -0
- package/knowledge-base/optimization/lagrangian-duality.en.md +50 -0
- package/knowledge-base/optimization/lagrangian-duality.md +50 -0
- package/knowledge-base/optimization/proximal-method.en.md +52 -0
- package/knowledge-base/optimization/proximal-method.md +52 -0
- package/knowledge-base/optimization/riemannian-optimization.en.md +51 -0
- package/knowledge-base/optimization/riemannian-optimization.md +51 -0
- package/knowledge-base/overview.en.md +29 -218
- package/knowledge-base/overview.md +29 -220
- package/knowledge-base/probability/concentration-inequality.en.md +45 -0
- package/knowledge-base/probability/concentration-inequality.md +45 -0
- package/knowledge-base/probability/entropy.en.md +47 -0
- package/knowledge-base/probability/entropy.md +47 -0
- package/knowledge-base/probability/fisher-information.en.md +48 -0
- package/knowledge-base/probability/fisher-information.md +48 -0
- package/knowledge-base/probability/information-bottleneck.en.md +49 -0
- package/knowledge-base/probability/information-bottleneck.md +49 -0
- package/knowledge-base/probability/kl-divergence.en.md +51 -0
- package/knowledge-base/probability/kl-divergence.md +51 -0
- package/knowledge-base/topology/euler-characteristic.en.md +54 -0
- package/knowledge-base/topology/euler-characteristic.md +54 -0
- package/knowledge-base/topology/fundamental-group.en.md +52 -0
- package/knowledge-base/topology/fundamental-group.md +52 -0
- package/knowledge-base/topology/persistent-homology.en.md +54 -0
- package/knowledge-base/topology/persistent-homology.md +54 -0
- package/lenses/algorithmic.en.md +49 -0
- package/lenses/algorithmic.md +49 -0
- package/lenses/axiomatization.en.md +48 -0
- package/lenses/axiomatization.md +48 -0
- package/lenses/categorical.en.md +48 -0
- package/lenses/categorical.md +48 -0
- package/lenses/causal.en.md +49 -0
- package/lenses/causal.md +49 -0
- package/lenses/duality.en.md +47 -0
- package/lenses/duality.md +47 -0
- package/lenses/game.en.md +50 -0
- package/lenses/game.md +50 -0
- package/lenses/geometric.en.md +53 -0
- package/lenses/geometric.md +53 -0
- package/lenses/local-to-global.en.md +52 -0
- package/lenses/local-to-global.md +52 -0
- package/lenses/perturbation.en.md +51 -0
- package/lenses/perturbation.md +51 -0
- package/lenses/probabilistic.en.md +48 -0
- package/lenses/probabilistic.md +48 -0
- package/lenses/projection.en.md +50 -0
- package/lenses/projection.md +50 -0
- package/lenses/spectral.en.md +52 -0
- package/lenses/spectral.md +52 -0
- package/lenses/symmetry.en.md +49 -0
- package/lenses/symmetry.md +49 -0
- package/lenses/topological.en.md +48 -0
- package/lenses/topological.md +48 -0
- package/lenses/variational.en.md +48 -0
- package/lenses/variational.md +48 -0
- package/package.json +9 -3
- package/references/books/abstract-algebra.md +5 -5
- package/references/books/algebraic-geometry-rising-sea.md +5 -5
- package/references/books/differential-geometry.md +3 -3
- package/references/books/matrix-analysis.md +4 -4
- package/references/books/micro-lie-theory.md +3 -3
- package/references/books/optimization-ml.md +1 -1
- package/references/books/smooth-manifolds.md +3 -3
- package/references/gpu-friendly-math.en.md +8 -8
- package/references/gpu-friendly-math.md +2 -2
- package/references/inspiration.en.md +1 -1
- package/references/inspiration.md +2 -2
- package/skills/math-research-activator/SKILL.en.md +161 -106
- package/skills/math-research-activator/SKILL.md +162 -110
- package/skills/math-research-activator/original-texts.en.md +9 -9
- package/skills/math-research-activator/original-texts.md +11 -11
- package/commands/abstraction.md +0 -11
- package/commands/algorithmic-thinking.md +0 -11
- package/commands/axiomatization.md +0 -11
- package/commands/causal-inference.md +0 -11
- package/commands/discrete-combinatorial.md +0 -11
- package/commands/game-theory.md +0 -11
- package/commands/induction-analogy.md +0 -11
- package/commands/information-theory.md +0 -11
- package/commands/logic-deduction.md +0 -11
- package/commands/modeling.md +0 -11
- package/commands/optimization.md +0 -11
- package/commands/probability-statistics.md +0 -11
- package/commands/symmetry-invariance.md +0 -11
- package/commands/topological-thinking.md +0 -11
- package/commands/transformation.md +0 -11
- package/skills/abstraction/SKILL.en.md +0 -117
- package/skills/abstraction/SKILL.md +0 -121
- package/skills/abstraction/original-texts.en.md +0 -163
- package/skills/abstraction/original-texts.md +0 -175
- package/skills/algorithmic-thinking/SKILL.en.md +0 -132
- package/skills/algorithmic-thinking/SKILL.md +0 -138
- package/skills/algorithmic-thinking/original-texts.en.md +0 -253
- package/skills/algorithmic-thinking/original-texts.md +0 -256
- package/skills/axiomatization/SKILL.en.md +0 -144
- package/skills/axiomatization/SKILL.md +0 -151
- package/skills/axiomatization/original-texts.en.md +0 -154
- package/skills/axiomatization/original-texts.md +0 -160
- package/skills/causal-inference/SKILL.en.md +0 -147
- package/skills/causal-inference/SKILL.md +0 -151
- package/skills/causal-inference/original-texts.en.md +0 -136
- package/skills/causal-inference/original-texts.md +0 -167
- package/skills/discrete-combinatorial/SKILL.en.md +0 -124
- package/skills/discrete-combinatorial/SKILL.md +0 -131
- package/skills/discrete-combinatorial/original-texts.en.md +0 -184
- package/skills/discrete-combinatorial/original-texts.md +0 -185
- package/skills/game-theory/SKILL.en.md +0 -117
- package/skills/game-theory/SKILL.md +0 -123
- package/skills/game-theory/original-texts.en.md +0 -131
- package/skills/game-theory/original-texts.md +0 -131
- package/skills/induction-analogy/SKILL.en.md +0 -145
- package/skills/induction-analogy/SKILL.md +0 -152
- package/skills/induction-analogy/original-texts.en.md +0 -140
- package/skills/induction-analogy/original-texts.md +0 -154
- package/skills/information-theory/SKILL.en.md +0 -134
- package/skills/information-theory/SKILL.md +0 -140
- package/skills/information-theory/original-texts.en.md +0 -127
- package/skills/information-theory/original-texts.md +0 -127
- package/skills/logic-deduction/SKILL.en.md +0 -130
- package/skills/logic-deduction/SKILL.md +0 -135
- package/skills/logic-deduction/original-texts.en.md +0 -160
- package/skills/logic-deduction/original-texts.md +0 -173
- package/skills/modeling/SKILL.en.md +0 -135
- package/skills/modeling/SKILL.md +0 -139
- package/skills/modeling/original-texts.en.md +0 -162
- package/skills/modeling/original-texts.md +0 -165
- package/skills/optimization/SKILL.en.md +0 -129
- package/skills/optimization/SKILL.md +0 -135
- package/skills/optimization/original-texts.en.md +0 -167
- package/skills/optimization/original-texts.md +0 -168
- package/skills/probability-statistics/SKILL.en.md +0 -146
- package/skills/probability-statistics/SKILL.md +0 -151
- package/skills/probability-statistics/original-texts.en.md +0 -191
- package/skills/probability-statistics/original-texts.md +0 -193
- package/skills/symmetry-invariance/SKILL.en.md +0 -135
- package/skills/symmetry-invariance/SKILL.md +0 -139
- package/skills/symmetry-invariance/original-texts.en.md +0 -206
- package/skills/symmetry-invariance/original-texts.md +0 -221
- package/skills/topological-thinking/SKILL.en.md +0 -124
- package/skills/topological-thinking/SKILL.md +0 -128
- package/skills/topological-thinking/original-texts.en.md +0 -134
- package/skills/topological-thinking/original-texts.md +0 -164
- package/skills/transformation/SKILL.en.md +0 -120
- package/skills/transformation/SKILL.md +0 -124
- package/skills/transformation/original-texts.en.md +0 -204
- package/skills/transformation/original-texts.md +0 -216
|
@@ -1,136 +1,188 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: math-research-activator
|
|
3
3
|
description: |
|
|
4
|
-
|
|
5
|
-
English:
|
|
4
|
+
数学研究操作系统:自动诊断用户意图,路由到思想透镜、数学知识库或设计翻译层。触发于设计/改进模型架构/算子/注意力、分析理论性质、迁移数学结构到 AI 设计。不触发于纯工程任务(debug、重构、调参)。
|
|
5
|
+
English: Mathematical research OS — auto-diagnoses user intent, routes to thinking lenses, math knowledge base, or design translation layer. Triggers on architecture/operator design, theoretical analysis, math-to-AI transfer. Does NOT trigger for pure engineering tasks.
|
|
6
6
|
---
|
|
7
7
|
|
|
8
|
-
> **语言路由**:若用户消息为英文,请读取并遵循同目录下的 `SKILL.en.md
|
|
8
|
+
> **语言路由**:若用户消息为英文,请读取并遵循同目录下的 `SKILL.en.md`;中文消息则继续使用本文件。
|
|
9
9
|
|
|
10
|
-
# 🧭
|
|
10
|
+
# 🧭 数学研究操作系统 / Math Research OS
|
|
11
11
|
|
|
12
|
-
> "
|
|
13
|
-
> "The model already holds enough mathematics inside; what's missing is a cross-domain activation. The human picks the direction; the agent searches, enumerates, verifies."
|
|
12
|
+
> "思想系统不负责给定理,知识系统不负责乱启发,设计层不负责装深刻。"
|
|
14
13
|
|
|
15
|
-
|
|
14
|
+
本系统是面向 AI 架构创新的数学参谋部——不是武器库,而是告诉你:**这场仗是什么仗、该用什么兵种、怎么部署、哪里会翻车。**
|
|
16
15
|
|
|
17
|
-
|
|
16
|
+
## 三层正交架构
|
|
18
17
|
|
|
19
|
-
|
|
20
|
-
|
|
18
|
+
| 层 | 职责 | 目录 | 核心问题 |
|
|
19
|
+
|----|------|------|---------|
|
|
20
|
+
| **思想透镜** | 诊断问题结构,推荐数学视角 | `../../lenses/*.md` | 这个问题该用什么视角看? |
|
|
21
|
+
| **数学知识** | 提供具体数学工具(定义/定理/公式) | `../../knowledge-base/*/*.md` | 这个视角需要哪些具体数学? |
|
|
22
|
+
| **设计翻译** | 把数学变成 AI 模块/loss/算子 | `../../design-patterns/*/*.md` | 这些数学怎么变成模型结构? |
|
|
21
23
|
|
|
22
|
-
|
|
24
|
+
辅助层:
|
|
25
|
+
- `../../references/books/*.md`:7 本书的蒸馏稿,需要深入时的完整上下文
|
|
26
|
+
- `../../references/gpu-friendly-math.md`:GPU 八维验收门(唯一权威)
|
|
27
|
+
- `../../agents/math-critic.md`:数学-工程双重批判器
|
|
23
28
|
|
|
24
|
-
##
|
|
29
|
+
## 意图诊断(5 场景)
|
|
25
30
|
|
|
26
|
-
|
|
31
|
+
| 场景 | 诊断信号 | 调用路径 |
|
|
32
|
+
|------|---------|---------|
|
|
33
|
+
| **A. 问题分析** | "这个设计合理吗?""逻辑链有没有漏洞?" | 透镜 → critic |
|
|
34
|
+
| **B. 机制设计** | "设计新 attention""把 X 迁移到 Y" | 透镜 → 知识 → 设计 → critic |
|
|
35
|
+
| **C. 知识查询** | "流形上的切空间是什么?""投影定理怎么用?" | 知识 |
|
|
36
|
+
| **D. 验证审查** | "这个公式成立吗?""loss 能保证什么?" | 知识 → 相关设计模式(若引用具体 AI 构造)→ critic |
|
|
37
|
+
| **E. 纯工程** | debug、重构、调参、代码审查 | **不调用数学系统** |
|
|
38
|
+
|
|
39
|
+
## 透镜库(15 个数学视角)
|
|
27
40
|
|
|
28
|
-
|
|
41
|
+
每个透镜回答:这是什么视角?适合诊断什么问题?会路由到哪些知识域?
|
|
42
|
+
|
|
43
|
+
| 透镜 | 文件 | 核心视角 |
|
|
44
|
+
|------|------|---------|
|
|
45
|
+
| 公理化 | `../../lenses/axiomatization.md` | 审查假设的相容性/独立性/完备性 |
|
|
46
|
+
| 对偶 | `../../lenses/duality.md` | 转换到对偶空间暴露约束与不变量 |
|
|
47
|
+
| 对称性 | `../../lenses/symmetry.md` | 变换下的不变量与守恒律 |
|
|
48
|
+
| 谱分解 | `../../lenses/spectral.md` | 特征值/奇异值揭示主导结构 |
|
|
49
|
+
| 几何 | `../../lenses/geometric.md` | 度量/曲率/流形上的空间结构 |
|
|
50
|
+
| 投影与分解 | `../../lenses/projection.md` | 正交分解、子空间分离、冲突消除 |
|
|
51
|
+
| 变分 | `../../lenses/variational.md` | 约束下极值、能量最小化 |
|
|
52
|
+
| 局部到整体 | `../../lenses/local-to-global.md` | 局部性质拼接为全局、层上同调障碍 |
|
|
53
|
+
| 拓扑 | `../../lenses/topological.md` | 连续变形不变量、连通性、空洞 |
|
|
54
|
+
| 范畴化 | `../../lenses/categorical.md` | 泛性质、函子、自然变换 |
|
|
55
|
+
| 扰动 | `../../lenses/perturbation.md` | 小扰动的传播、稳定性、鲁棒性 |
|
|
56
|
+
| 因果 | `../../lenses/causal.md` | 相关≠因果、干预、反事实 |
|
|
57
|
+
| 博弈 | `../../lenses/game.md` | 多方策略互动、均衡、机制设计 |
|
|
58
|
+
| 概率统计 | `../../lenses/probabilistic.md` | 量化不确定性、贝叶斯更新 |
|
|
59
|
+
| 算法 | `../../lenses/algorithmic.md` | 复杂度、可行性、并行性 |
|
|
60
|
+
|
|
61
|
+
## 知识库(按数学领域组织)
|
|
62
|
+
|
|
63
|
+
每个知识卡片回答:最小定义、核心公式、适用问题、AI 设计翻译、工程可行性、风险。
|
|
64
|
+
|
|
65
|
+
| 领域 | 目录 | 知识卡片 |
|
|
66
|
+
|------|------|---------|
|
|
67
|
+
| 矩阵分析 | `../../knowledge-base/matrix-analysis/` | projection, spectral-decomposition, low-rank-approximation, positive-semidefinite, matrix-perturbation |
|
|
68
|
+
| 最优化 | `../../knowledge-base/optimization/` | lagrangian-duality, convex-optimization, constrained-optimization, riemannian-optimization, proximal-method |
|
|
69
|
+
| 微分几何 | `../../knowledge-base/differential-geometry/` | manifold, tangent-space, metric-tensor, geodesic, curvature, connection |
|
|
70
|
+
| 李理论 | `../../knowledge-base/lie-theory/` | group-action, lie-group, lie-algebra, representation, equivariance |
|
|
71
|
+
| 拓扑 | `../../knowledge-base/topology/` | persistent-homology, euler-characteristic, fundamental-group |
|
|
72
|
+
| 概率与信息 | `../../knowledge-base/probability/` | concentration-inequality, entropy, kl-divergence, information-bottleneck, fisher-information |
|
|
73
|
+
| 信息几何 | `../../knowledge-base/information-geometry/` | natural-gradient, fisher-metric |
|
|
74
|
+
|
|
75
|
+
## 设计模式库(按 AI 组件组织)
|
|
76
|
+
|
|
77
|
+
每个设计模式回答:数学来源、AI 模块形式、可实现结构、GPU 可行性、论文表述、风险。
|
|
78
|
+
|
|
79
|
+
| 组件类型 | 目录 | 模式 |
|
|
80
|
+
|---------|------|------|
|
|
81
|
+
| 注意力 | `../../design-patterns/attention/` | projection-attention, spectral-attention, equivariant-attention, geometry-aware-attention, information-bottleneck-attention |
|
|
82
|
+
| 损失函数 | `../../design-patterns/loss/` | orthogonality-loss, contrastive-loss, variational-loss, information-bottleneck-loss, constraint-penalty |
|
|
83
|
+
| 路由 | `../../design-patterns/routing/` | optimal-transport-routing, graph-routing, moe-routing, spectral-clustering-routing |
|
|
84
|
+
| 表示 | `../../design-patterns/representation/` | shared-private-decomposition, manifold-representation, equivariant-split, subspace-alignment |
|
|
85
|
+
| 压缩 | `../../design-patterns/compression/` | low-rank-kv-cache, spectral-token-pruning, topology-preserving-compression, leverage-score-selection |
|
|
86
|
+
|
|
87
|
+
## 自动触发条件
|
|
88
|
+
|
|
89
|
+
**必须同时满足 Gate 1 + Gate 2 + Gate 3 才介入:**
|
|
90
|
+
|
|
91
|
+
### Gate 0 · 排除门(最高优先级)
|
|
92
|
+
以下任务**无论工作区含什么**都不触发:代码审查、debug、重构、调参、构建部署、纯事实查询、通用软件工程。
|
|
93
|
+
|
|
94
|
+
### Gate 1 · 环境信号
|
|
95
|
+
工作区含架构核心代码(attention/transformer/MoE、`*.cu`/kernel)或研究笔记。仅 `model.py`、`trainer.py` 等常规文件**不构成**环境信号。
|
|
96
|
+
|
|
97
|
+
### Gate 2 · 任务信号
|
|
98
|
+
用户任务涉及**设计/改进**新架构/算子、**分析**理论性质、**迁移**数学结构到 AI 设计,或**查询与 AI 研究相关的数学知识**(如"切空间在优化中怎么用")。纯百科式数学查询(如"什么是群"且无 AI 上下文)不自动触发,但可通过 `/ask` 手动进入。
|
|
99
|
+
|
|
100
|
+
### Gate 3 · 意图匹配
|
|
101
|
+
用户意图匹配场景 A/B/C/D 之一。纯工程任务匹配场景 E → 不介入。
|
|
102
|
+
|
|
103
|
+
> **`/ask` 入口**:手动调用时跳过 Gate 1 和 Gate 2,仅执行 Gate 0(排除门)+ Gate 3(意图匹配),可直接进入任意场景包括知识查询。
|
|
104
|
+
|
|
105
|
+
## 主流程
|
|
106
|
+
|
|
107
|
+
### 第一步:诊断意图
|
|
108
|
+
1. 判断用户意图属于场景 A/B/C/D/E 哪个
|
|
109
|
+
2. 提取问题核心张力:想保留什么?想抑制什么?约束是什么?工程瓶颈是什么?
|
|
110
|
+
3. 输出问题类型分类
|
|
111
|
+
|
|
112
|
+
### 第二步:路由调用
|
|
113
|
+
|
|
114
|
+
```
|
|
115
|
+
场景 A(分析):选 1-3 个透镜 → 输出视角诊断 → critic 审查
|
|
116
|
+
场景 B(设计):选 1-3 个透镜 → 查知识卡片 → 生成设计模式 → critic 审查
|
|
117
|
+
场景 C(查询):直接加载知识卡片 → 按知识激活协议输出
|
|
118
|
+
场景 D(验证):加载知识卡片 → critic 审查条件与边界
|
|
119
|
+
场景 E(工程):不介入
|
|
120
|
+
```
|
|
121
|
+
|
|
122
|
+
### 第三步:输出格式
|
|
123
|
+
|
|
124
|
+
**场景 A/B 输出**:
|
|
125
|
+
1. **[诊断]** 问题类型 + 核心张力
|
|
126
|
+
2. **[透镜]** 推荐 1-3 个数学视角(标注为什么适合/不适合)
|
|
127
|
+
3. **[知识]**(仅场景 B)需要的具体数学工具(引用知识卡片)
|
|
128
|
+
4. **[设计]**(仅场景 B)候选 AI 模块草案(引用设计模式)
|
|
129
|
+
5. **[GPU]** 候选过八维门(友好/可改造/不友好)
|
|
130
|
+
6. **[结论]** 保留通过双验收门的候选 + 下一步建议
|
|
131
|
+
|
|
132
|
+
**场景 C 输出**(知识激活协议):
|
|
133
|
+
1. 最小定义
|
|
134
|
+
2. 核心公式
|
|
135
|
+
3. 适用问题
|
|
136
|
+
4. AI 设计翻译
|
|
137
|
+
5. 工程可行性
|
|
138
|
+
6. 风险与失效条件
|
|
139
|
+
7. 深入参考(书蒸馏稿 / 原书路径)
|
|
140
|
+
|
|
141
|
+
**场景 D 输出**:
|
|
142
|
+
1. 成立条件
|
|
143
|
+
2. 不成立条件
|
|
144
|
+
3. 最多能保证什么
|
|
145
|
+
4. 不能保证什么
|
|
146
|
+
5. 工程可行性
|
|
29
147
|
|
|
30
|
-
|
|
31
|
-
- 代码审查 / debug / 参数传递链核查 / 接口一致性检查
|
|
32
|
-
- 重构、重命名、删除冗余代码、清理死代码
|
|
33
|
-
- 构建 / 打包 / CI / 部署 / 环境配置
|
|
34
|
-
- 纯事实查询("这个函数干什么"、"这个参数传给了谁")
|
|
35
|
-
- 通用软件工程(文件 I/O、网络、数据加载、日志)
|
|
36
|
-
- 训练脚本调参、超参搜索、实验对比
|
|
37
|
-
- 添加 / 修改 loss 项的实现细节(非设计新 loss 的数学结构)
|
|
38
|
-
|
|
39
|
-
**判断口诀:如果任务可以用"阅读代码 → 追踪调用链 → 报告结果"完成,就不需要数学武器。**
|
|
40
|
-
|
|
41
|
-
### Gate 1 · 环境信号(必要条件,非充分条件)
|
|
42
|
-
|
|
43
|
-
工作区含以下至少一类:
|
|
44
|
-
- 模型架构核心代码(非训练脚本/数据管道):attention/transformer/MoE 实现、`*.cu`/`*.cuh`/kernel、Triton/CUDA 算子。
|
|
45
|
-
- 算法研究笔记 / 论文审查文档。
|
|
46
|
-
- 新架构 / 新算子的设计稿或数学推导。
|
|
47
|
-
|
|
48
|
-
> 仅有 `model.py`、`trainer.py`、`config.json` 等常规 ML 工程文件**不构成环境信号**。
|
|
49
|
-
|
|
50
|
-
### Gate 2 · 任务信号(必要条件,非充分条件)
|
|
51
|
-
|
|
52
|
-
用户当前任务明确涉及以下至少一项:
|
|
53
|
-
- **设计或改进**一个新的模型架构 / 算子 / 注意力机制(非修复已有实现)。
|
|
54
|
-
- **选择或论证**数学结构的适用性(如"该不该用流形约束"、"这个结构有没有等变性")。
|
|
55
|
-
- **分析**算法的理论性质(复杂度下界、收敛性、表达力、信息瓶颈)。
|
|
56
|
-
- 把某个数学领域的结构**迁移**到算法/GPU 设计中。
|
|
57
|
-
|
|
58
|
-
> 如果任务信号仅为"代码跑不通"、"参数没传过去"、"loss 没生效"等工程问题,即使环境信号命中也**不触发**。
|
|
59
|
-
|
|
60
|
-
**不介入 / When NOT to use:**
|
|
61
|
-
- Gate 0 命中(排除门)。
|
|
62
|
-
- Gate 1 或 Gate 2 任一未命中。
|
|
63
|
-
- 问题不属于数学能帮助的范畴。
|
|
64
|
-
|
|
65
|
-
## 主流程 / The Activation Loop
|
|
66
|
-
|
|
67
|
-
> 详细工作方式见 `../../references/agentic-workflow.md`(Human-in-the-Agent-Loop)。
|
|
68
|
-
|
|
69
|
-
1. **诊断 / Diagnose**:算法结构或瓶颈是什么?(复杂度?显存/KV?数值?并行?表达力?)
|
|
70
|
-
2. **映射 / Map**:用「现代数学工具箱」(下)扫描可迁移的结构,**枚举多个候选**(发挥大上下文优势,别只给一个)。
|
|
71
|
-
3. **路由 / Route**:选 1–3 个思想武器深入(决策树见下)。
|
|
72
|
-
4. **GPU 筛选 / Screen**:每个候选过 `../../references/gpu-friendly-math.md` 八维,给「友好/可改造/不友好」+ 改造建议。
|
|
73
|
-
5. **双验收门 / Gate**:只保留**数学正确 AND(八维友好或可改造)**的候选。
|
|
74
|
-
6. **追踪 / Track**:复杂探索用 markdown testplan 表(模板见 agentic-workflow.md)迭代收敛。
|
|
75
|
-
|
|
76
|
-
八维正式术语必须保持一致:**张量化 / GEMM 可映射 / 复杂度 / 显存与 KV-Cache / 低精度稳定 / 并行与通信 / 稀疏结构 / 算子融合**。不要用只覆盖部分维度的模糊判断替代八维门。
|
|
77
|
-
|
|
78
|
-
## 思想武器路由决策树 / Weapon Routing
|
|
79
|
-
|
|
80
|
-
按问题核心特征匹配(最多选 3 个,标主/辅):
|
|
81
|
-
|
|
82
|
-
1. **多方互动**(我的最优取决于他人)→ `/game-theory`(主);涉资源分配+`/optimization`;信息不对称+`/information-theory`
|
|
83
|
-
2. **不确定性/随机性** → `/probability-statistics`(主);需因果而非相关+`/causal-inference`
|
|
84
|
-
3. **约束下求最优** → `/optimization`(主);需先建模+`/modeling`(前置)
|
|
85
|
-
4. **当前形式难处理,需换视角/化简** → `/transformation`(主)
|
|
86
|
-
5. **需提取本质结构** → `/abstraction`(主);验证假设+`/axiomatization`;简化+`/symmetry-invariance`
|
|
87
|
-
6. **需严格推理验证** → `/logic-deduction`(主);验证前提+`/axiomatization`
|
|
88
|
-
7. **从数据/经验找规律** → `/induction-analogy`(主);跨域迁移+`/abstraction`
|
|
89
|
-
8. **构建预测/解释模型** → `/modeling`(主);优化+`/optimization`;不确定+`/probability-statistics`
|
|
90
|
-
9. **变化中的不变性/守恒/等变** → `/symmetry-invariance`(主);连通结构+`/topological-thinking`
|
|
91
|
-
10. **化为可执行步骤/评估可行性与复杂度** → `/algorithmic-thinking`(主)
|
|
92
|
-
11. **压缩/编码/信息瓶颈/KV-Cache 压缩/量化** → `/information-theory`(主);涉及表示变换+`/transformation`;涉及路由信息增益+`/game-theory`
|
|
93
|
-
12. **有限对象的计数/枚举/结构** → `/discrete-combinatorial`(主)
|
|
94
|
-
|
|
95
|
-
> **现代数学优先提示**:当问题是「设计/改进算子或结构」时,路由武器的同时**务必先开现代数学工具箱**——很多突破来自把代数几何/微分几何/李理论的结构迁移过来,而非只在经典工具里打转。
|
|
96
|
-
|
|
97
|
-
## 现代数学工具箱(Layer 3 · 按需加载)/ Modern-Math Toolbox
|
|
148
|
+
**必须给出结论,不得只输出分析而不收敛。**
|
|
98
149
|
|
|
99
|
-
|
|
150
|
+
## GPU 八维验收门
|
|
100
151
|
|
|
101
|
-
|
|
102
|
-
|
|
103
|
-
| 算子=矩阵乘/谱/低秩/数值稳定 | `matrix-analysis.md` | GEMM 化、低秩压缩、谱归一化、预条件 |
|
|
104
|
-
| 训练/收敛/优化器/约束 | `optimization-ml.md` | 自适应/二阶优化的可行性、对偶求解 |
|
|
105
|
-
| 对称/等变/半环/置换不变 | `abstract-algebra.md` | 群等变层、热带半环路由、有限域编码 |
|
|
106
|
-
| 流形约束/隐空间几何/可微结构 | `smooth-manifolds.md` | 流形优化、Stiefel/正交约束、测地插值 |
|
|
107
|
-
| 度量/曲率/自然梯度/规范/纤维丛 | `differential-geometry.md` | 自然梯度/K-FAC、信息几何、gauge 等变 |
|
|
108
|
-
| 位姿/SO(3)/SE(3)/状态估计/等变 | `micro-lie-theory.md` | 李群优化、SE(3) 等变、流形损失 |
|
|
109
|
-
| 注意力/稀疏/全局一致性/KV 压缩 | `algebraic-geometry-rising-sea.md` | sheaf 注意力、Čech 上同调正则、Plücker KV、热带门控 |
|
|
152
|
+
正式术语(唯一权威来源:`../../references/gpu-friendly-math.md`):
|
|
153
|
+
**张量化 / GEMM 可映射 / 复杂度 / 显存与 KV-Cache / 低精度稳定 / 并行与通信 / 稀疏结构 / 算子融合**
|
|
110
154
|
|
|
111
|
-
##
|
|
155
|
+
## 深度查阅协议
|
|
112
156
|
|
|
113
|
-
-
|
|
114
|
-
-
|
|
115
|
-
-
|
|
157
|
+
- **轻度**:读知识卡片(`../../knowledge-base/*/*.md`),自足可用
|
|
158
|
+
- **中度**:读书蒸馏稿(`../../references/books/*.md`),获取更完整上下文
|
|
159
|
+
- **深度**:本机有 `math_book/<PDF>` 时,Agent 自动 `pdftotext` + grep 定位原文页
|
|
116
160
|
|
|
117
|
-
##
|
|
161
|
+
## 工作流范例
|
|
118
162
|
|
|
119
|
-
|
|
163
|
+
**用户**:"设计新的 KV Cache 压缩方法,保留长期依赖,不想只做 top-k"
|
|
120
164
|
|
|
121
|
-
|
|
165
|
+
```
|
|
166
|
+
第一步 诊断:场景 B(机制设计)
|
|
167
|
+
问题类型:序列记忆压缩 + 信息保留 + 长程结构
|
|
168
|
+
核心张力:压缩 token 数量 vs 不破坏长期依赖
|
|
122
169
|
|
|
123
|
-
|
|
124
|
-
1.
|
|
125
|
-
2.
|
|
126
|
-
3.
|
|
127
|
-
4. **[GPU 筛选]**:每个候选过八维,给「友好/可改造/不友好」+ 改造建议。
|
|
128
|
-
5. **[结论]**:保留同时通过双验收门的候选;必要时给 testplan 表。
|
|
170
|
+
第二步 透镜选择:
|
|
171
|
+
1. 谱分解(保留主导子空间)
|
|
172
|
+
2. 信息论(保留最大互信息状态)
|
|
173
|
+
3. 拓扑(保留序列结构关键连接点)
|
|
129
174
|
|
|
130
|
-
|
|
175
|
+
第三步 知识查询:
|
|
176
|
+
→ low-rank-approximation(矩阵分析)
|
|
177
|
+
→ leverage-score-selection(矩阵分析)
|
|
178
|
+
→ information-bottleneck(概率与信息)
|
|
131
179
|
|
|
132
|
-
|
|
180
|
+
第四步 设计翻译:
|
|
181
|
+
候选 A:Spectral KV Compression(低秩 + leverage score)
|
|
182
|
+
候选 B:Information-Preserving Cache(query sensitivity)
|
|
183
|
+
候选 C:Topology-Preserving Cache(图桥接节点保留)
|
|
133
184
|
|
|
134
|
-
|
|
135
|
-
|
|
136
|
-
|
|
185
|
+
第五步 Critic 审查:
|
|
186
|
+
A 最 GPU 友好,B 需估计未来 query 有不确定性,C 图构建成本过高
|
|
187
|
+
建议:优先 A,B 作轻量 gate
|
|
188
|
+
```
|
|
@@ -1,4 +1,4 @@
|
|
|
1
|
-
# Intellectual Sources for the
|
|
1
|
+
# Intellectual Sources for the Math Research OS
|
|
2
2
|
|
|
3
3
|
## The Tradition of Problem Classification and Tool Selection
|
|
4
4
|
|
|
@@ -22,7 +22,7 @@ Pólya's central insight: **different types of problems require different proble
|
|
|
22
22
|
|
|
23
23
|
> "Problem solving is a practical skill, like swimming — you can learn it by imitation and practice." — Pólya
|
|
24
24
|
|
|
25
|
-
Pólya's taxonomy of heuristics provides the intellectual foundation for our
|
|
25
|
+
Pólya's taxonomy of heuristics provides the intellectual foundation for our activator: **the core characteristics of a problem (such as "does it involve uncertain quantities," "does it exhibit symmetry," "can it be decomposed into subproblems") determine the most appropriate problem-solving method**.
|
|
26
26
|
|
|
27
27
|
### Newell & Simon's Problem Space Theory
|
|
28
28
|
|
|
@@ -42,10 +42,10 @@ Alan H. Schoenfeld, in *Mathematical Problem Solving* (1985), analyzed why stude
|
|
|
42
42
|
|
|
43
43
|
- **Resources**: Knowledge, skills, tools — corresponding to our 15 intellectual weapons
|
|
44
44
|
- **Heuristics**: How to use resources — corresponding to the methodological workflow for each intellectual weapon
|
|
45
|
-
- **Control**: When to use which resource — corresponding to the function of the
|
|
45
|
+
- **Control**: When to use which resource — corresponding to the function of the activator
|
|
46
46
|
- **Belief Systems**: Beliefs about mathematics and about oneself — affecting whether the appropriate tool is selected
|
|
47
47
|
|
|
48
|
-
Core insight: **mastery of tools ≠ ability to solve problems. The critical difference lies at the 'control' level — knowing when to deploy which tool**. This is precisely the problem the
|
|
48
|
+
Core insight: **mastery of tools ≠ ability to solve problems. The critical difference lies at the 'control' level — knowing when to deploy which tool**. This is precisely the problem the activator is designed to address.
|
|
49
49
|
|
|
50
50
|
> "The students' problem is not a lack of knowledge or skill, but a lack of strategic decision-making ability — they do not know when to use which method." — Schoenfeld
|
|
51
51
|
|
|
@@ -65,7 +65,7 @@ Characteristics of wicked problems:
|
|
|
65
65
|
- Solutions are not right or wrong, only better or worse
|
|
66
66
|
- Span multiple domains
|
|
67
67
|
|
|
68
|
-
Core insight: **wicked problems require combinations of cross-domain intellectual weapons, whereas tame problems can be solved with a single tool**. The
|
|
68
|
+
Core insight: **wicked problems require combinations of cross-domain intellectual weapons, whereas tame problems can be solved with a single tool**. The activator recommends multi-tool combinations when facing wicked problems and single-tool focus when facing tame problems.
|
|
69
69
|
|
|
70
70
|
> "Wicked problems have no right or wrong answers, only better or worse ways of dealing with them." — Rittel & Webber
|
|
71
71
|
|
|
@@ -76,7 +76,7 @@ Daniel Kahneman, in *Thinking, Fast and Slow* (2011), proposed the dual-system t
|
|
|
76
76
|
- **System 1**: Fast, intuitive, automatic — responsible for the majority of everyday decisions
|
|
77
77
|
- **System 2**: Slow, rational, deliberate — required for decisions demanding careful thought
|
|
78
78
|
|
|
79
|
-
Core insight: **not all problems require System 2-level rational analysis**. Many everyday decisions require only intuition (System 1), and forcibly deploying intellectual weapons constitutes over-analysis. The
|
|
79
|
+
Core insight: **not all problems require System 2-level rational analysis**. Many everyday decisions require only intuition (System 1), and forcibly deploying intellectual weapons constitutes over-analysis. The activator's "inapplicable scenarios" list is grounded in this insight — simple problems do not need tools; wicked problems do.
|
|
80
80
|
|
|
81
81
|
> "Overthinking is the enemy of decision-making — some decisions are best left to intuition." — Kahneman
|
|
82
82
|
|
|
@@ -87,15 +87,15 @@ Imre Lakatos, in *Proofs and Refutations* (1976), demonstrated that mathematical
|
|
|
87
87
|
- **Progressive Problemshift**: Each revision gives rise to new, deeper questions
|
|
88
88
|
- **Degenerative Problemshift**: Revisions merely accommodate counterexamples without generating new insights
|
|
89
89
|
|
|
90
|
-
Core insight: **tool selection itself should be a progressive problemshift** — the initial choice may not be ideal, but through post-use reflection, the ability to choose improves progressively. The
|
|
90
|
+
Core insight: **tool selection itself should be a progressive problemshift** — the initial choice may not be ideal, but through post-use reflection, the ability to choose improves progressively. The activator's "combination sequencing" and "auxiliary perspectives" recommendations embody this dialectical thinking.
|
|
91
91
|
|
|
92
92
|
> "Knowledge is not truth descending from above, but conjecture continually refined through criticism and refutation." — Lakatos
|
|
93
93
|
|
|
94
94
|
## Mapping Ideas to Practice
|
|
95
95
|
|
|
96
|
-
Our
|
|
96
|
+
Our activator integrates all of the above intellectual traditions:
|
|
97
97
|
|
|
98
|
-
| Intellectual Source | Manifestation in the
|
|
98
|
+
| Intellectual Source | Manifestation in the Math Research OS |
|
|
99
99
|
|---------|-------------------|
|
|
100
100
|
| Pólya's Heuristics | 11 characteristic branches in the decision tree — each branch corresponds to a class of heuristic strategies |
|
|
101
101
|
| Newell & Simon's Problem Space | Problem characteristic dimensions — interactivity, uncertainty, constraint, structure, dynamism |
|
|
@@ -1,4 +1,4 @@
|
|
|
1
|
-
#
|
|
1
|
+
# 数学研究操作系统的思想来源 / Intellectual Sources for the Math Research OS
|
|
2
2
|
|
|
3
3
|
## 问题分类与工具选择的传统 / The Tradition of Problem Classification and Tool Selection
|
|
4
4
|
|
|
@@ -22,7 +22,7 @@ Pólya 的核心洞见:**不同类型的问题需要不同的解题策略**。
|
|
|
22
22
|
|
|
23
23
|
> "解题是一种实践技能,就像游泳一样——你可以通过模仿和练习来学习它。" —— Pólya
|
|
24
24
|
|
|
25
|
-
Pólya
|
|
25
|
+
Pólya 的启发法分类为我们的数学研究操作系统提供了思想基础:**问题的核心特征(如"是否涉及不确定量""是否有对称性""能否分解为子问题")决定了最合适的解题方法**。
|
|
26
26
|
|
|
27
27
|
### Newell & Simon 的问题空间理论 / Newell & Simon's Problem Space Theory
|
|
28
28
|
|
|
@@ -40,12 +40,12 @@ Allen Newell 和 Herbert A. Simon 在《人类问题求解》(Human Problem Solv
|
|
|
40
40
|
|
|
41
41
|
Alan H. Schoenfeld 在《数学问题求解》(Mathematical Problem Solving, 1985) 中分析了为什么学生即使掌握了解题工具也无法解决问题——关键不是缺少工具,而是缺少**策略选择能力**:
|
|
42
42
|
|
|
43
|
-
- **资源** (Resources):知识、技能、工具——相当于我们的15
|
|
44
|
-
- **启发法** (Heuristics)
|
|
45
|
-
- **控制** (Control)
|
|
43
|
+
- **资源** (Resources):知识、技能、工具——相当于我们的15 个思想透镜
|
|
44
|
+
- **启发法** (Heuristics):如何使用资源——相当于每种思想透镜的方法流程
|
|
45
|
+
- **控制** (Control):何时使用哪个资源——相当于数学研究操作系统的功能
|
|
46
46
|
- **信念系统** (Belief Systems):对数学和自我的信念——影响是否选择合适的工具
|
|
47
47
|
|
|
48
|
-
核心洞见:**掌握工具≠能解决问题。关键差异在于'控制'
|
|
48
|
+
核心洞见:**掌握工具≠能解决问题。关键差异在于'控制'层面——知道何时调用哪个工具**。这正是数学研究操作系统要解决的问题。
|
|
49
49
|
|
|
50
50
|
> "学生的问题不是缺乏知识或技能,而是缺乏策略性决策能力——他们不知道何时该用什么方法。" —— Schoenfeld
|
|
51
51
|
|
|
@@ -65,7 +65,7 @@ Horst Rittel 和 Melvin Webber 在 "Dilemmas in a General Theory of Planning" (1
|
|
|
65
65
|
- 解没有对错之分,只有好与更好
|
|
66
66
|
- 跨越多个领域
|
|
67
67
|
|
|
68
|
-
|
|
68
|
+
核心洞见:**棘问题需要跨领域的思想透镜组合,而驯问题可以用单一工具解决**。数学研究操作系统在面对棘问题时推荐多工具组合,面对驯问题时推荐单一聚焦。
|
|
69
69
|
|
|
70
70
|
> "棘问题没有对错答案,只有更好或更差的处理方式。" —— Rittel & Webber
|
|
71
71
|
|
|
@@ -76,7 +76,7 @@ Daniel Kahneman 在《思考,快与慢》(Thinking, Fast and Slow, 2011) 中
|
|
|
76
76
|
- **系统1** (System 1):快速、直觉、自动——日常生活中大部分决策
|
|
77
77
|
- **系统2** (System 2):慢速、理性、可控——需要深思熟虑的决策
|
|
78
78
|
|
|
79
|
-
核心洞见:**不是所有问题都需要系统2级别的理性分析**。生活中的许多决策只需要直觉(系统1
|
|
79
|
+
核心洞见:**不是所有问题都需要系统2级别的理性分析**。生活中的许多决策只需要直觉(系统1),强行调用思想透镜是过度分析。数学研究操作系统的"不适用场景"清单正是基于这个洞见——简单问题不需要工具,棘问题才需要。
|
|
80
80
|
|
|
81
81
|
> "过度思考是决策的敌人——有些决定最好交给直觉。" —— Kahneman
|
|
82
82
|
|
|
@@ -87,15 +87,15 @@ Imre Lakatos 在《证明与反驳》(Proofs and Refutations, 1976) 中展示了
|
|
|
87
87
|
- **进步的问题转换** (Progressive Problemshift):每次修正产生新的、更深刻的问题
|
|
88
88
|
- **退化的问题转换** (Degenerative Problemshift):修正只是为了应付反例,没有产生新洞见
|
|
89
89
|
|
|
90
|
-
|
|
90
|
+
核心洞见:**工具选择本身也应该是一个进步的问题转换**——第一次选择可能不理想,但通过使用后反思,选择能力逐步提升。数学研究操作系统的"组合顺序"和"辅助视角"推荐正是这种辩证思维的体现。
|
|
91
91
|
|
|
92
92
|
> "知识不是从天而降的真理,而是在批评与反驳中不断改进的猜想。" —— Lakatos
|
|
93
93
|
|
|
94
94
|
## 从思想到实践的映射 / Mapping Ideas to Practice
|
|
95
95
|
|
|
96
|
-
|
|
96
|
+
我们的数学研究操作系统融合了以上思想:
|
|
97
97
|
|
|
98
|
-
| 思想来源 |
|
|
98
|
+
| 思想来源 | 在数学研究操作系统中的体现 |
|
|
99
99
|
|---------|-------------------|
|
|
100
100
|
| Pólya 启发法 | 决策树中的11个特征分支——每个分支对应一类启发策略 |
|
|
101
101
|
| Newell & Simon 问题空间 | 问题特征维度刻画——互动性、不确定性、约束性、结构性、动态性 |
|
package/commands/abstraction.md
DELETED
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: abstraction
|
|
3
|
-
description: |
|
|
4
|
-
从复杂问题提取数学结构、发现跨领域共性、构建通用理论,或为算法/算子设计寻找可迁移的抽象结构时调用。路由到 ../skills/abstraction/SKILL.md。
|
|
5
|
-
English: Trigger when extracting mathematical structures, finding cross-domain commonalities, building general theories, or finding transferable abstract structures for algorithm/operator design. Routes to ../skills/abstraction/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/abstraction/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/abstraction/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: algorithmic-thinking
|
|
3
|
-
description: |
|
|
4
|
-
问题需有限步骤求解、做算法设计与复杂度/可行性/并行性分析、证明终止性/正确性,或为 GPU 设计可执行算法/算子时调用。路由到 ../skills/algorithmic-thinking/SKILL.md。
|
|
5
|
-
English: Trigger when a problem needs finite-step solving, algorithm design with complexity/tractability/parallelism analysis, termination/correctness proofs, or designing executable algorithms/operators for GPU. Routes to ../skills/algorithmic-thinking/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/algorithmic-thinking/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/algorithmic-thinking/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: axiomatization
|
|
3
|
-
description: |
|
|
4
|
-
审查假设合理性、构建公理系统、发现理论矛盾、验证相容性/独立性/完备性,或为算法/算子定公理与不变量并检验可计算性时调用。路由到 ../skills/axiomatization/SKILL.md。
|
|
5
|
-
English: Trigger when auditing assumptions, building axiom systems, discovering contradictions, verifying consistency/independence/completeness, or defining axioms/invariants for algorithms/operators and checking computability. Routes to ../skills/axiomatization/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/axiomatization/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/axiomatization/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: causal-inference
|
|
3
|
-
description: |
|
|
4
|
-
处理相关≠因果、干预/反事实/do-演算/因果图建模/混淆变量识别/效应评估,或为可解释性/分布外泛化/DGP 建模引入显式因果假设时调用。路由到 ../skills/causal-inference/SKILL.md。
|
|
5
|
-
English: Trigger when handling correlation≠causation, intervention/counterfactual/do-calculus/causal DAG/confounder/effect estimation, or introducing explicit causal assumptions for interpretability/OOD generalization/DGP modeling. Routes to ../skills/causal-inference/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/causal-inference/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/causal-inference/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: discrete-combinatorial
|
|
3
|
-
description: |
|
|
4
|
-
需计数/枚举/发现有限结构规律、处理图论/组合结构/生成函数/递推关系,或为稀疏/路由/拓扑结构设计组合方案时调用。路由到 ../skills/discrete-combinatorial/SKILL.md。
|
|
5
|
-
English: Trigger when needing to count/enumerate/find patterns in finite structures, handling graph theory/combinatorial structures/generating functions/recurrences, or designing combinatorial schemes for sparse/routing/topological structures. Routes to ../skills/discrete-combinatorial/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/discrete-combinatorial/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/discrete-combinatorial/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
package/commands/game-theory.md
DELETED
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: game-theory
|
|
3
|
-
description: |
|
|
4
|
-
策略取决于他人选择、需计算纳什均衡/分析博弈/做机制设计,或为多智能体/对抗训练/路由博弈设计策略时调用。路由到 ../skills/game-theory/SKILL.md。
|
|
5
|
-
English: Trigger when optimal strategy depends on others' choices, needing Nash equilibrium/game analysis/mechanism design, or designing strategies for multi-agent/adversarial training/routing games. Routes to ../skills/game-theory/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/game-theory/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/game-theory/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: induction-analogy
|
|
3
|
-
description: |
|
|
4
|
-
从数据找规律、做数学归纳证明(弱/强/结构/超限)、跨域类比迁移结构,或为算法/算子设计借鉴其他领域结构时调用。路由到 ../skills/induction-analogy/SKILL.md。
|
|
5
|
-
English: Trigger when finding patterns from data, doing mathematical induction proofs (weak/strong/structural/transfinite), transferring structures across domains by analogy, or borrowing structures from other fields for algorithm/operator design. Routes to ../skills/induction-analogy/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/induction-analogy/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/induction-analogy/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: information-theory
|
|
3
|
-
description: |
|
|
4
|
-
以"信息=不确定性的减少"度量信息价值、计算熵/互信息/KL 散度/信道容量,或为压缩/量化/路由设计信息准则时调用。路由到 ../skills/information-theory/SKILL.md。
|
|
5
|
-
English: Trigger when measuring information value as "information = reduction of uncertainty," computing entropy/mutual information/KL divergence/channel capacity, or designing information criteria for compression/quantization/routing. Routes to ../skills/information-theory/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/information-theory/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/information-theory/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: logic-deduction
|
|
3
|
-
description: |
|
|
4
|
-
检查证明严谨性、做形式/谓词逻辑分析、发现逻辑漏洞,或为算法正确性与不变量做形式推导与证明验证时调用。路由到 ../skills/logic-deduction/SKILL.md。
|
|
5
|
-
English: Trigger when checking proof rigor, doing formal/predicate logic analysis, finding logical loopholes, or doing formal derivation and proof verification for algorithm correctness and invariants. Routes to ../skills/logic-deduction/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/logic-deduction/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/logic-deduction/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
package/commands/modeling.md
DELETED
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: modeling
|
|
3
|
-
description: |
|
|
4
|
-
将现实问题转化为数学问题(现实→数学→解释)、构建预测/解释模型、做量纲分析与模型选择,或为算法/算子建立可计算模型时调用。路由到 ../skills/modeling/SKILL.md。
|
|
5
|
-
English: Trigger when translating real-world problems into math, building predictive/explanatory models, doing dimensional analysis and model selection, or building computable models for algorithms/operators. Routes to ../skills/modeling/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/modeling/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/modeling/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
package/commands/optimization.md
DELETED
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: optimization
|
|
3
|
-
description: |
|
|
4
|
-
问题涉及约束下取舍与目标极值、需用拉格朗日/KKT/对偶分析,或为算法/算子/训练设计选择优化方法时调用。路由到 ../skills/optimization/SKILL.md。
|
|
5
|
-
English: Trigger when a problem involves trade-offs and extrema under constraints, needs Lagrangian/KKT/duality analysis, or chooses optimization methods for algorithm/operator/training design. Routes to ../skills/optimization/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/optimization/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/optimization/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: probability-statistics
|
|
3
|
-
description: |
|
|
4
|
-
量化不确定性、做贝叶斯推断/假设检验/回归建模/实验设计,或为随机算法/采样/量化/训练动力学设计概率机制时调用。路由到 ../skills/probability-statistics/SKILL.md。
|
|
5
|
-
English: Trigger when quantifying uncertainty, doing Bayesian inference/hypothesis testing/regression/experimental design, or designing probabilistic mechanisms for random algorithms/sampling/quantization/training dynamics. Routes to ../skills/probability-statistics/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/probability-statistics/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/probability-statistics/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|
|
@@ -1,11 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: symmetry-invariance
|
|
3
|
-
description: |
|
|
4
|
-
做群论/不变量/Noether 定理/对称性破缺分析,或为算法设计等变性/守恒量/不变结构时调用。路由到 ../skills/symmetry-invariance/SKILL.md。
|
|
5
|
-
English: Trigger when doing group theory/invariant/Noether's theorem/symmetry-breaking analysis, or designing equivariance/conservation/invariant structures for algorithms. Routes to ../skills/symmetry-invariance/SKILL.md.
|
|
6
|
-
---
|
|
7
|
-
|
|
8
|
-
若用户消息为英文,读取并遵循 `../skills/symmetry-invariance/SKILL.en.md`,按其操作规程输出;否则读取并遵循 `../skills/symmetry-invariance/SKILL.md`,按其操作规程输出(用于算法/算子/GPU 设计时含 [GPU 可行性] 项)。
|
|
9
|
-
|
|
10
|
-
当前问题:
|
|
11
|
-
$ARGUMENTS
|