Long-form Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
31.15ROUGE-L
57
Jun 2, 2026
27.13ROUGE-L
54
Feb 27, 2026
87Precision
29
Apr 30, 2026
63.5Token-F1
29
Apr 30, 2026
78.65VeriScore F1@k
26
Jul 7, 2026
81.7VeriScore F1@k
26
Jul 7, 2026
61.87VeriScore F1@k
26
Jul 7, 2026
70.11VeriScore F1@k
26
Jul 7, 2026
88.93VeriScore F1@k
26
Jul 7, 2026
90.33VeriScore F1@k
26
Jul 7, 2026
82.54VeriScore F1@k
26
Jul 7, 2026
91.59VeriScore F1@k
26
Jul 7, 2026
85.3LLM-Acc
20
Feb 26, 2026
93.1LLM Accuracy
20
Feb 26, 2026
91.18str-em
19
Apr 10, 2026
23.34ROUGE-L
18
Apr 21, 2026
49.7VeriScore F1
14
Feb 26, 2026
66.9VeriScore F1
14
Feb 26, 2026
61VeriScore F1
14
Feb 26, 2026
75.9VeriScore F1
14
Feb 26, 2026
61.7Win Rate vs. Holistic Reward
13
Feb 26, 2026
31.5F1
11
Feb 27, 2026
78VeriScore F1@k
10
Jul 7, 2026
83.43VeriScore F1@k
10
Jul 7, 2026
87.56VeriScore F1@k
10
Jul 7, 2026