NExT-QA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
NExT-QA open
61LLM Judge Score
4
NExT-QA original (test)
67Temporal Score
4
NExT-QA de-biased set
60.8Temporal Reasoning Score
4
NEXT-QA (test)
30.3WUPS
4
NExT-QA Hard Split - Causal
56.4Accuracy
3
NExT-QA Hard Split - Temporal
49.8Accuracy
3
NExT-QA (OOD)
72.63CH (Accuracy)
2