LLM Wiki — compiled memory
https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f
Three-layer directory: raw/ (immutable sources), wiki/ (LLM-maintained compiled pages), CLAUDE.md (schema). LLM reads sources and integrates them into the wiki — updating pages, resolving contradictions — so knowledge compounds across sessions instead of being retrieved cold.
At a glance
- Type
- Compilation over RAG
- Tier
- T5
- Created
- see post date
- Latest release
- not applicable — informal artifact (gist/tweet/blog)
- License
- not applicable — informal artifact (no formal license)
- Pricing
- not applicable — theoretical / informal
- Funding
- not applicable — not commercial
Taxonomy
- storage
- file
- retrieval
- injection
- persistence
- long-term
- update
- overwrite
- unit
- file
- governance
- editable
- conflict
- n/a
When to use
Optimised for: not applicable - theoretical / not a system
Anti-fit: not applicable - theoretical / not a system
Pros & cons
Pros
Three-layer compilation paradigm (raw / wiki / schema) directly challenges RAG orthodoxy — knowledge compounds across sessions instead of being retrieved cold.
Cons
GitHub Gist; conceptual sketch with minimal implementation guidance.
Claims & capabilities
Karpathy, April 2026 GitHub Gist. Challenges RAG orthodoxy; separates ingestion (compilation) from retrieval.
Technical surface
- API surface
- not applicable — theoretical / not a system
- Backend storage
- not applicable — theoretical / not a system
- Deployment
- not applicable — theoretical / informal
- Embedding model
- not applicable — theoretical / not a system
- Multi-tenancy
- not applicable — theoretical / not a system
- MCP
- not applicable — theoretical / informal idea
- A2A
- not applicable — theoretical / informal idea
- OpenTelemetry
- not applicable — theoretical / informal idea
Similar systems
Other theoretical / informal — ideas without a paper in the catalog, ranked by inbound references.
- From Human Memory to AI Memory (survey) T4
Eight-quadrant classification grid across personal/system, parametric/non-parametric, and short-term/long-term axes. Bridges cognitive-science memory taxonomy to LLM architecture choices, less common than purely engineering-oriented surveys. v2 revision April 23, 2025.
- Context Engineering T5
"+1 for 'context engineering' over 'prompt engineering' … the delicate art and science of filling the context window with just the right information for the next step." The LLM is "a coworker with anterograde amnesia" — cannot consolidate or build long-running knowledge once training ends.
- Context Engineering — naming event T5
Endorses "context engineering" as a distinct discipline from prompt engineering — what agents do with their context window (routing, compression, tool output formatting, memory retrieval injection) is engineering, not just prompting.
- Context Expansion Law T5
"Application context tends to expand to fill the context limits supported by the model." Treats agent memory as a first-class unresolved design problem rather than a solved component; explicitly defers a memory deep-dive to a future post.
- Externalization in LLM Agents T4
Traces the shift from weights-as-capability to harness-as-capability; analyzes memory, skills, and protocols as three coupled forms of externalization and examines how they interact. Memory is defined as the externalization of state across time. Covers self-evolving harnesses and shared agent infrastructure as emerging directions. April 9, 2026.
- Files Are All You Need T5
Coding agents (Claude Code, Cursor) converge on the filesystem as their primary memory abstraction: conversation histories as searchable files, skills as files, retrieval via file search rather than vector DBs. Argument: LLMs are fluent with filesystem concepts, so the filesystem is the right interface even if storage underneath is a database.