Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

GEO-AT

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multimodal ReasoningGEO-AT RNA
METEOR0.491
17
Multimodal ReasoningGEO-AT DNA
METEOR52.9
17
Multimodal ReasoningGEO-AT Protein
METEOR41.7
17
Multimodal ReasoningGEO-AT Molecule
METEOR0.415
17
Functional group hallucination testGEO-AT Proteins (test)
HR10
3
Functional group hallucination testGEO-AT Molecules (test)
HR23
3
Showing 6 of 6 rows