ResearchBenchmarksMulti-task Driving Scene Understanding Robustness on The Dolphins (Lvl. 2)Follow44.97Final ScoreNutVLM36.109238.409640.7143.0104Feb 9, 2026Evaluation ResultsMethodMethodLinksFinal ScoreNutVLM2026.0244.97NRP2026.0244.8AAA2026.0244.45Bit-Red2026.0244.1MS2026.0243.75Scene–CADA2026.0243.58JPEG2026.0241.5TVM2026.0236.45