LongBench
Benchmarks
TaskDatasetSOTA resultTrendResults
LongBench
58.4Average Score
369
LongBench
60.31M-Avg
294
LongBench V2
82.36Overall Score
185
LongBench (test)
58.7Avg Score
166
LongBench (test)
51.87Average Score
147
LongBench
34F1 Score
143
LongBench
62.1Overall Average Score
143
LongBench OOD v2
53QA
120
LongBench v2
68.9Average Score
113
LongBench 1.0 (test)
32.94NarrativeQA
108
LongBench
73.8Score
107
LongBench
70.4Accuracy (LongBench)
101
LongBench
31.96Average Score
96
LongBench (test)
71.4FewShot Performance
94
LongBench-e
53.04Average Score
93
LongBench
58.4Average Score
90
LongBench
57.15HotpotQA
82
LongBench
53.67MultifieldQA Score
75
LongBench 1.0 (test)
61.5MultiNews
73
LongBench (test)
7,011HotpotQA
69
LongBench Qasper
0.446F1
62
LongBench v2
46.32Overall Accuracy
62
LongBench
49.8Aggregate Score
60
LongBench
103Accuracy
60
LongBench
59.71HotPotQA Accuracy
59