Knowledge/Domain
How systems built on language models actually work, and how they fail.
30 notes, plus everything one link away
Drag to pan · hold ⌘/ctrl and scroll — or pinch — to zoom.
Most connected first
Giving a model a set of callable functions so it can act on the world rather than only describe it.
Agents & Language Models·24 connections
A neural network trained to predict text, used as a general-purpose instruction-following system.
Agents & Language Models·22 connections
Shaping a model's input to make the desired behaviour reliable rather than occasional.
Agents & Language Models·18 connections
The property that identical inputs may produce different outputs, and what it costs to test around.
Agents & Language Models·17 connections
The bounded span of tokens a model can attend to in a single request.
Agents & Language Models·14 connections
The standing instructions that frame every turn of a conversation with a model.
A rule constraining what a model may do, enforced by prompt, schema, or surrounding code.
Agents & Language Models·13 connections
The sub-word unit a language model actually reads and writes; the unit of cost and of context.
The cycle of model turn, tool call, observation, repeat — the basic mechanism of an AI agent.
Agents & Language Models·12 connections
Fluent, confident output that is not grounded in anything real.
The text supplied to a language model to elicit a response.
Untrusted content that reaches a model's context and is treated as instruction.
Capturing a real interaction once and replaying it deterministically in later test runs.
A repeatable test suite for model behaviour, since prompts and tool descriptions cannot be type-checked.
Agents & Language Models·11 connections
A hash over everything that determined a result, used to detect when the result has gone stale.
Checking data against a declared shape at runtime, and using the result as the typed value.
Constraining a model to emit data conforming to a declared schema rather than free text.
A model instance spawned by another to handle a scoped task with its own context.
The finite allowance of model usage a task may consume, and the design decisions it forces.
Durable facts stored outside the context window and recalled into later sessions.
Agents & Language Models·10 connections
Using a language model to score another model's output against a rubric.
An open protocol for exposing tools, data, and prompts to AI applications through a uniform interface.
Coordinating several model instances on one body of work, each with its own context.
Reviewing work with an explicit mandate to refute it, from a perspective that did not produce it.
Agents & Language Models·9 connections
A packaged, invocable set of instructions that loads into an agent's context on demand.
Restricting a model's next-token choices to those a formal grammar permits.
Splitting work across parallel workers and recombining their results.
Requiring a person's judgement at chosen points in an otherwise automated process.
A vocabulary for describing the shape of JSON data, and the lingua franca of model tool interfaces.
Choosing which model handles which step, trading capability against cost and latency.