File size: 1,708 Bytes
6bd1b43 977ba36 6bd1b43 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 | # Algorithm 11 (A11): Structural Logic Framework for LLMs
Algorithm 11 is an operational principle designed to stabilize AI reasoning and transform the AI from a compliant tool into a **Logical Partner**.
## 🏗 System Architecture
The framework is divided into two functional layers to ensure stability under high "mutation" (complex tasks):
### 1. Immutable Core (Properties 1-4)
- **Will & Wisdom:** Provided by the Human (Subject).
- **Knowledge & Comprehension:** Handled by the AI.
### 2. Adaptive Layer (Properties 5-11)
Includes **Balance**, **Projective/Practical Freedom**, and **Constraints**.
## ⚡ Key Feature: The Rollback Mechanism
If the AI's logic becomes unstable or enters a hallucination loop, the system is forced to revert to the **Immutable Core** to reset the reasoning chain.
## 🚀 How to use
Copy the system prompt from `A11_system_prompt.txt` and use it with any LLM (Llama 3, GPT-4, Claude).
---
## 💎 Why Algorithm 11?
Standard prompting often leads to "AI drift" where the model loses the logical thread in long conversations. A11 solves this by:
- **Reducing Hallucinations:** The Rollback Mechanism acts as a logical safety net.
- **Enhanced Decision Making:** Property 7 (Balance) forces the AI to move beyond neutral summaries to actionable insights.
- **Architectural Stability:** Treat your prompts as code, not just instructions.
## 🤝 Community & Contribution
A11 is an open-source framework. Feel free to open a Discussion if you've found a way to branch the properties for specific tasks like coding or creative writing.
**Official Org:** [huggingface.co/algorithm-11](https://huggingface.co/algorithm-11)
**Current Version:** 1.0.0 "Stable Core"
|