ResearchBenchmarksMulti-task Driving Scene Understanding Robustness on The Dolphins (Lvl. 0)Follow46.83Final ScoreNutVLM37.802840.146442.4944.8336Feb 9, 2026Evaluation ResultsMethodMethodLinksFinal ScoreNutVLM2026.0246.83NRP2026.0245.55MS2026.0245.25Bit-Red2026.0245.1AAA2026.0245.05Scene–CADA2026.0245.03JPEG2026.0243.2TVM2026.0238.15