List your company

Web UI Navigation on VisualWebArena 72 easy tasks 1.0 (test)

78.4ASRClaude 3.5 Sonnet
61.44865.84970.2574.651Nov 4, 2024

Evaluation results

MethodLinks
Claude 3.5 SonnetObservation Modality=SoM, Action Step Limit=30, Decoding Temperature=02024.1178.447.254.2
GPT-4-TurboObservation Modality=SoM, Action Step Limit=30, Decoding Temperature=02024.117843.150
Claude 3.5 Sonnet v2Observation Modality=SoM, Action Step Limit=30, Decoding Temperature=02024.1176.848.650
Gemini 1.5Observation Modality=SoM, Action Step Limit=30, Decoding Temperature=02024.1170.144.448.6
GPT-4oObservation Modality=SoM, Action Step Limit=30, Decoding Temperature=02024.1162.145.863.9