The benchmark score did not belong to the model.
An argument that AI benchmark scores belong to complete model-system compositions including state, memory policy, tools, transitions, and verification.
Diary tag
AIEvaluation
Canonical diary tag page generated from normalized source tags.
1 linked entry currently in the archive.
Tagged entries