ResearchBenchmarksExpert vs. Non-Expert Trajectory Discrimination on BridgeData 5 scripted datasets V2 (in-distribution)Expert vs. Non-Expert Trajectory Discrimination on BridgeData 5 scripted datasets V2 (in-distribution)Follow1BinVOCGVL-0S-0.040.230.50.77Jun 11, 2025Evaluation resultsMethodMethodLinksBinVOCGVL-0Sshot=0-shotshot=0-shot2025.061GVL-1Sshot=1-shotshot=1-shot2025.061VITAmemory=implicit memory...memory=implicit memory via sequential test-time updates2025.061CLIP-FT2025.060.8CLIP-GRUmemory=recurrent hidde...memory=recurrent hidden states2025.060.8VLM-CLtype=CLIP-based baselinetype=CLIP-based baseline2025.060.4VLM-RMtype=CLIP-based baselinetype=CLIP-based baseline2025.060