Something went wrong. Try again.
about things notes.zzstoatzz.io
Something went wrong. Try again.
README.md
ai #
building with and around models, organized by concept: the models themselves, the programs around them, and how they read and remember.
contents #
- decision-guided-decoding — decision probabilities as text policies, candidate-set normalization, and how end-token weighting changes reply length
- prompt-caching — the prefix invariant, per-model floors, write/read economics, and measuring hit rate on anthropic and openai
- retrieval/ — query-time mechanics: asymmetric embedding, rank fusion, synthesis between retrieval and prompt
- memory/ — how agents keep state between runs: message archives, deliberate vs background writes, write-time curation
- models/ — specific models, decision models among them, and what a swap changes
- local-models/ — running models on your own hardware: serving, tool-calling, GPU runtimes
- harnesses/ — the program around a model: adapters, background work in pi, terminal control, skill discovery
- mcp/ — the model context protocol: servers, clients, oauth, skills
- projected-embedding-clusters — reading structure off an embedded corpus: umap, hdbscan, work items from geometry
- named-entity-recognition — what NER returns, why decoder confidence measures the label rather than the span, and what actually removes noise
- reimplementing-inference — porting a model to another runtime: the feature extraction drifts, not the weights, and how to measure it
- computer-use — what a driver can act on depends on what the app exposes; where the time goes; keeping the chooser bounded
- customer-facing-agent-ui — a product agent needs its own frontend and a proxy that authorizes thread ownership; LangSmith Studio and Remote MCP are operator surfaces
- audio-model-evaluation — testing whether a model's descriptions of music follow the audio
adjacent, filed elsewhere on purpose: storage/turbopuffer (vector-store operations are storage operations), and the agent-memory design notes that live in the bot repo's docs beside the code they describe.