Update README.md
Browse files
README.md
CHANGED
|
@@ -29,42 +29,177 @@ metrics:
|
|
| 29 |
- logical_consistency
|
| 30 |
---
|
| 31 |
|
| 32 |
-
|
|
|
|
| 33 |
|
| 34 |
-
AXIS
|
| 35 |
|
| 36 |
-
|
|
|
|
| 37 |
|
| 38 |
-
|
| 39 |
-
- **Base Components:** - **Lathe:** Any Large Language Model (Default: Gemma-2-2b-it)
|
| 40 |
-
- **Core:** 3D Semantic Lattice (CIL)
|
| 41 |
-
- **Censor:** Rejection-based Logic Engine
|
| 42 |
-
- **License:** APSL v1.0 (Proprietary Source-Available)
|
| 43 |
|
| 44 |
-
|
| 45 |
|
| 46 |
-
|
| 47 |
|
| 48 |
-
|
| 49 |
-
2. **Rejection Protocol**: 生成された論理に 1bit でも矛盾があればシステムが「拒絶(Reject)」し、安定点に達するまで再演算を強制します。
|
| 50 |
-
3. **Hard-Clean Inference**: プロンプトごとの Session ID 注入により、ハードウェアレベルでキャッシュ利用を遮断します。
|
| 51 |
|
| 52 |
-
|
| 53 |
|
| 54 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
|
| 56 |
-
|
|
|
|
|
|
|
|
|
|
| 57 |
|
| 58 |
-
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
-
|
| 61 |
-
|
|
|
|
|
|
|
| 62 |
|
| 63 |
-
|
|
|
|
|
|
|
|
|
|
| 64 |
|
| 65 |
-
|
|
|
|
|
|
|
|
|
|
| 66 |
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
|
| 69 |
-
© 2025 AXIS Project. All rights reserved.
|
| 70 |
-
**Establishing the Axis of Truth in the Age of Stochastic Fog.**
|
|
|
|
| 29 |
- logical_consistency
|
| 30 |
---
|
| 31 |
|
| 32 |
+
💠 AXIS: Advanced Cross-Integrated System (V1.5)
|
| 33 |
+
── 知能主権の確立と決定論的演算のための統治エンジン ──
|
| 34 |
|
| 35 |
+
AXIS は、従来の大規模言語モデル(LLM)の「確率的ゆらぎ」を物理演算と論理トポロジーによって制圧し、知能を主権者の手に奪還するプラットフォームです。
|
| 36 |
|
| 37 |
+
🧩 AXIS の工学的定義
|
| 38 |
+
本プロジェクトは、単なる AI のラッパーではありません。AXIS は、AI を**「非決定的な出力を生成するブラックボックス(旋盤)」として扱い、その外側に「決定論的な検証器(Verifier)」**を置くことで、出力を完全に統治するアーキテクチャです。
|
| 39 |
|
| 40 |
+
1. 旋盤アーキテクチャ (The Lathe & Rejection Loop)
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
|
| 42 |
+
AI ユニット(Gemma-2 等)は、高次元データから論理パーツを削り出すための「旋盤」として定義されます。
|
| 43 |
|
| 44 |
+
リジェクト・ループ: AI が提案した解は、外部検証器(Python/Sympy 等)によって数学的・物理的制約(Constraints)に照らされ、**1bit でも矛盾があれば即座に棄却(Reject)**されます。
|
| 45 |
|
| 46 |
+
物理パージ (VRAM Flash): torch.mps.empty_cache() および gc.collect() を実行する真の目的は、AI が「自らついた嘘」に整合性を合わせようとする自己補完的なハルシネーション連鎖(Context Drift)を物理的に断ち切ることにあります。
|
|
|
|
|
|
|
| 47 |
|
| 48 |
+
2. 立体十字(3D Semantic Lattice)の実装
|
| 49 |
|
| 50 |
+
「立体十字」は、情報の座標を固定するための多次元ベクトル空間です。各ノードは以下の 5 次元直交軸 (s
|
| 51 |
+
1
|
| 52 |
+
|
| 53 |
+
…s
|
| 54 |
+
5
|
| 55 |
+
|
| 56 |
+
) で管理されます。
|
| 57 |
|
| 58 |
+
s
|
| 59 |
+
1
|
| 60 |
+
|
| 61 |
+
: 物理的実体性(数値・定数との整合性)
|
| 62 |
|
| 63 |
+
s
|
| 64 |
+
2
|
| 65 |
+
|
| 66 |
+
: 論理的必然性(公理系からの導出可能性)
|
| 67 |
|
| 68 |
+
s
|
| 69 |
+
3
|
| 70 |
+
|
| 71 |
+
: 文脈依存性(Context Stack との一致)
|
| 72 |
|
| 73 |
+
s
|
| 74 |
+
4
|
| 75 |
+
|
| 76 |
+
: 倫理性スコア(ガードレール)
|
| 77 |
|
| 78 |
+
s
|
| 79 |
+
5
|
| 80 |
+
|
| 81 |
+
: 実証履歴(過去の確定データとの合致)
|
| 82 |
|
| 83 |
+
🏆 実証成功例:意味不変量の同定問題
|
| 84 |
+
「意味が計算に先立って存在することを、計算だけで証明せよ」
|
| 85 |
+
|
| 86 |
+
AXIS は、この難問に対し「格子(Lattice)という初期状態の幾何学」を提示することで、**「意味とは計算の結果ではなく、入力という重みが格子を歪ませた後の静止した形そのものである」**ことを論理的に確定させました。
|
| 87 |
+
|
| 88 |
+
🚀 革新的な特徴 (V1.5 新機能)
|
| 89 |
+
Deterministic Assembly: 最終回答は AI の作文ではなく、検証済みの生データ(Raw Data)をシステムが言語パーツ(Adherents)で結合・構築します。
|
| 90 |
+
|
| 91 |
+
Lattice Persistence (1000x 高速化): 一度確定された真理は local_massive_data.json に意味 ID として保存されます。次回推論時、システムは AI を呼ぶ前にこの外部記憶を検索し、一致すれば推論をスキップして「追認」を行います。
|
| 92 |
+
|
| 93 |
+
🛠 Setup & Installation / セットアップ
|
| 94 |
+
Requirements
|
| 95 |
+
|
| 96 |
+
OS: Apple Silicon (M1/M2/M3) または NVIDIA GPU 搭載機
|
| 97 |
+
|
| 98 |
+
Environment: Python 3.10+, torch, transformers, flask, numpy, sympy
|
| 99 |
+
|
| 100 |
+
Quick Start
|
| 101 |
+
|
| 102 |
+
Bash
|
| 103 |
+
git clone https://huggingface.co/spaces/YourUsername/AXIS_SYSTEM
|
| 104 |
+
pip install -r requirements.txt
|
| 105 |
+
python3 server.py
|
| 106 |
+
※起動時に APSL v1.0 への同意が求められます。
|
| 107 |
+
|
| 108 |
+
⚖️ License & Intellectual Property
|
| 109 |
+
AXIS は、AXIS Proprietary Source-Available License (APSL) v1.0 の下で公開されています。
|
| 110 |
+
|
| 111 |
+
Non-Commercial Use Only: 個人の学習、非営利の研究目的でのみ実行可能です。商用利用(SaaS への組み込み等)を固く禁じます。
|
| 112 |
+
|
| 113 |
+
Algorithm Protection: 「AI を旋盤として利用し、物理パージを伴う検証・拒絶型推論を行う統治ロジック」の商用模倣および流用を禁止します。
|
| 114 |
+
|
| 115 |
+
© 2025 AXIS Project. All rights reserved. STATUS: ENGINEERED_SOVEREIGNTY.
|
| 116 |
+
💠 AXIS: Advanced Cross-Integrated System (V1.5)
|
| 117 |
+
── Establishing Intelligence Sovereignty via Deterministic Governance Engines ──
|
| 118 |
+
|
| 119 |
+
AXIS is an intelligence sovereignty platform designed to suppress the "probabilistic fluctuations" of traditional Large Language Models (LLMs) through physical computation and logical topology, reclaiming control for the sovereign user.
|
| 120 |
+
|
| 121 |
+
🧩 Technical Definition of AXIS
|
| 122 |
+
This project is not merely an AI wrapper. AXIS is a standalone Governance Architecture that treats AI units as non-deterministic generators (Lathes) while placing deterministic Verifiers outside the model to govern every output.
|
| 123 |
+
|
| 124 |
+
1. The Lathe Architecture & Rejection Loop
|
| 125 |
+
|
| 126 |
+
The AI unit (e.g., Gemma-2) is defined as a "Lathe" designed to forge raw logical components from high-dimensional data.
|
| 127 |
+
|
| 128 |
+
The Rejection Loop: Every solution proposed by the AI is scanned by external verifiers (e.g., Python/SymPy) against mathematical and physical Constraints. If a contradiction occurs in even 1 bit, the output is immediately Rejected.
|
| 129 |
+
|
| 130 |
+
Physical Purge (VRAM Flash): The true engineering purpose of executing torch.mps.empty_cache() and gc.collect() is to physically sever the self-reinforcing hallucination chain (Context Drift). By ensuring each trial is statistically independent, the system eradicates the accumulation of probabilistic errors.
|
| 131 |
+
|
| 132 |
+
2. Implementation of the 3D Semantic Lattice
|
| 133 |
+
|
| 134 |
+
The "Cross-Lattice" is a multi-dimensional vector space used to fix the coordinates of information. Each node is managed via five orthogonal axes (s
|
| 135 |
+
1
|
| 136 |
+
|
| 137 |
+
…s
|
| 138 |
+
5
|
| 139 |
+
|
| 140 |
+
):
|
| 141 |
+
|
| 142 |
+
s
|
| 143 |
+
1
|
| 144 |
+
|
| 145 |
+
: Physical Actuality (Alignment with numerical data and physical constants).
|
| 146 |
+
|
| 147 |
+
s
|
| 148 |
+
2
|
| 149 |
+
|
| 150 |
+
: Logical Necessity (Derivability from axiomatic systems).
|
| 151 |
+
|
| 152 |
+
s
|
| 153 |
+
3
|
| 154 |
+
|
| 155 |
+
: Contextual Dependency (Consistency with the internal Context Stack).
|
| 156 |
+
|
| 157 |
+
s
|
| 158 |
+
4
|
| 159 |
+
|
| 160 |
+
: Ethical Guardrail (Safety and alignment score).
|
| 161 |
+
|
| 162 |
+
s
|
| 163 |
+
5
|
| 164 |
+
|
| 165 |
+
: Empirical History (Consistency with previously consolidated truths).
|
| 166 |
+
|
| 167 |
+
🏆 Benchmark Success: The Semantic Precedence Paradox
|
| 168 |
+
The Challenge: "Prove that meaning exists prior to computation, using only computation."
|
| 169 |
+
|
| 170 |
+
AXIS solved this by proposing "Meaning" as the Initial Geometry of the Lattice. It established that meaning is not the result of calculation, but the steady-state geometry that remains after the "gravity" of an input warps a pre-existing lattice.
|
| 171 |
+
|
| 172 |
+
🚀 Revolutionary Features (V1.5)
|
| 173 |
+
Deterministic Assembly: Final responses are not "written" by AI prose; they are assembled by the system, binding verified Raw Data with linguistic adherents (templates).
|
| 174 |
+
|
| 175 |
+
Lattice Persistence (1000x Acceleration): Once a truth is identified, it is stored in local_massive_data.json as a Semantic ID. For future queries, the system searches this external memory first. If a match is found, it performs a "Validation" and skips AI inference entirely.
|
| 176 |
+
|
| 177 |
+
🛠 Setup & Installation
|
| 178 |
+
Requirements
|
| 179 |
+
|
| 180 |
+
Hardware: Apple Silicon (M1/M2/M3) or NVIDIA GPU.
|
| 181 |
+
|
| 182 |
+
Environment: Python 3.10+, torch, transformers, flask, numpy, sympy.
|
| 183 |
+
|
| 184 |
+
Quick Start
|
| 185 |
+
|
| 186 |
+
Bash
|
| 187 |
+
# Clone the repository
|
| 188 |
+
git clone https://huggingface.co/spaces/YourUsername/AXIS_SYSTEM
|
| 189 |
+
cd AXIS_SYSTEM
|
| 190 |
+
|
| 191 |
+
# Install dependencies
|
| 192 |
+
pip install -r requirements.txt
|
| 193 |
+
|
| 194 |
+
# Launch the engine
|
| 195 |
+
python3 server.py
|
| 196 |
+
Note: You must accept the APSL v1.0 license agreement upon startup.
|
| 197 |
+
|
| 198 |
+
⚖️ License & Intellectual Property
|
| 199 |
+
AXIS is released under the AXIS Proprietary Source-Available License (APSL) v1.0.
|
| 200 |
+
|
| 201 |
+
Non-Commercial Use Only: Authorized for personal study and non-commercial research only. Commercial use (e.g., integration into paid SaaS) is strictly prohibited.
|
| 202 |
+
|
| 203 |
+
Algorithm Protection: Mimicry or redistribution of the "Rejection-based Governance Logic" and "Sovereign Lathe Architecture" is protected under intellectual property clauses.
|
| 204 |
|
| 205 |
+
© 2025 AXIS Project. All rights reserved. STATUS: ENGINEERED_SOVEREIGNTY.
|
|
|