Concept

Capture of an LLM at inference time through system prompts, fine-tunes, distillations, refusal policies, and tool-use constraints. The cheapest of the LLM capture surfaces because it changes nothing about the model — same weights, used differently per request — and per the capture-taxonomy a surface capture rather than a consumer-key capture. Most opaque in practice because the configurations are typically not published and can change silently between requests.

Connections

This is a graph landing page. Use the local graph and backlinks (right) to explore, or open the interactive ontology.