AI Architecture
- LLM Architecture Layers Client → API Wrapper → Guardrails → Inference Engine → Weights → Output Filter GitHub Pages (html) Cloudflare Pages (md)
- LLM Physical Deliverables What files make up a model GitHub Pages (html) Cloudflare Pages (md)
- AI Chatbot Full Workflow From your input to response GitHub Pages (html) Cloudflare Pages (md)
- Switch Provider Which layers change when switching provider GitHub Pages (html) Cloudflare Pages (md)
- Contributors Who does what GitHub Pages (html) Cloudflare Pages (md)
Compression & AGI
- Compression = Intelligence = World Model Shannon limit, DeepMind paper, Brain-as-Mini-Nature theory GitHub Pages (html) Cloudflare Pages (md)
- Compression Demo — Huffman Coding & Typos Concrete bitstream demo, prefix codes, typo penalty GitHub Pages (html) Cloudflare Pages (md)
- LoRA Fine-Tuning — Customising LLMs How LoRA updates 0.1% params without breaking original quality GitHub Pages (html) Cloudflare Pages (md)
- Identifying Unseen Concepts in LLMs How to tell if a concept exists in the model's knowledge space GitHub Pages (html) Cloudflare Pages (md)
- LLM Security — Shared Context & Prompt Injection Two attack vectors, enterprise deployment risks, prevention GitHub Pages (html) Cloudflare Pages (md)
- LLM Encryption & Privacy TEE, homomorphic encryption, private encoding (Navajo approach) GitHub Pages (html) Cloudflare Pages (md)