VL_Correctness

Test Models

YOLO (vision)
ResNet  (vision)
Clip (vision-language)
Florence-2 (vision-language)
LLaVA (vision-language)
LLaMA (language)
Gemini-2 (language)
Gemma (language)

Methodologies

Comparison of embeddings using cloud of points (Dario), mathematical distance (E.g., Euclidian, Cosine, Siameses; Seongha)

Data

Cloud (Waggle, Sage)
Wild fire (HPWREN, Konza)
Wild life (Konza)
Sound (Morton Arboretum)