Skip to content

Evaluation

WidgetCaps evaluation methodology note

issues found in at least three previous works' evaluations of WidgetCaps, rendering numerical comparisons invalid

Multimodal Google DeepMind — Gemma

How it was run

Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.

PaliGemma 1
Google DeepMind — Gemma

partial detail 13 method fields published

See how this test was run