List your company

Temporal Semantic Understanding on Temporal Semantics Vinoground and TimeBlind (test)

70.9Task 1 AccuracyUnifiedReward-Flex
-2.62816.46135.5554.639Jun 10, 2026

Evaluation results

MethodLinks
UnifiedReward-FlexMethod Category=Reasoning-based Reward Model, Evaluation Mode=pairwise2026.0670.93723.1
UnifiedReward-ThinkMethod Category=Reasoning-based Reward Model, Evaluation Mode=pairwise2026.066738.319.8
UnifiedReward2.0Method Category=Non-reasoning Reward Model2026.0658.852.530.8
VideoScore1.1Method Category=Non-reasoning Reward Model2026.0651.347.21.3
SG-PVRMethod Category=Reasoning-based Reward Model, Evaluation Mode=pointwise2026.0651.351.931.4
OmniScoreMethod Category=Training-free Baseline2026.0647.753.86
VideoScore2Method Category=Reasoning-based Reward Model2026.064240.416.1
ETVAMethod Category=Training-free Baseline2026.0633.33111.2
VBench2.0Method Category=Training-free Baseline2026.0622.421.610.2
VisionRewardMethod Category=Non-reasoning Reward Model2026.0620.2190.5
VideoRewardMethod Category=Non-reasoning Reward Model2026.0619.2170.8
LiFTMethod Category=Non-reasoning Reward Model2026.060.20.20