Skip to content

Evaluation

TaxCollector23 edited this page Oct 10, 2026 · 2 revisions

Evaluation

Evaluation runners are not part of the current shipped CLI. AXIOM can help prepare the inputs around an AI job, but it does not claim to execute a benchmark or publish quality results.

When an evaluation command is added, document its input contract, output format, seed, model provenance, and measurement method here before calling it available.

Clone this wiki locally