ResearchBenchmarksMultimodal Understanding on Cross-LVLM Aggregate (test)Follow13.8ECEBICR12.70820.07927.4534.821May 11, 2026Evaluation ResultsMethodMethodLinksECEBSAccuracyF1 ScoreAUCPRAUROCBICR2026.0513.821.466.857.964.268.1II2026.051825.159.561.659.563PIK2026.0521.325.759.462.964.868.4SAPLMA2026.0522.927.859.663.761.165.7CCPS2026.0528.734.354.558.158.559.2PE2026.0535.635.547.960.152.156.3Self-Probing2026.054040.351.360.958.459.8P(True)2026.0541.139.851.752.861.665.6