Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
58.17LM Score
9
Apr 7, 2026
63.41LM Score
9
Apr 7, 2026
60.62F1 Score
9
Apr 7, 2026
43.62F1
9
Apr 7, 2026
0.86Normalized Latency (vs Vanilla RAG)
9
Apr 1, 2026
0.83Normalized Average Latency
9
Apr 1, 2026
0.8Average Latency (Normalized)
9
Apr 1, 2026
0.88Average Latency (Normalized)
9
Apr 1, 2026
0.85Normalized Average Latency
9
Apr 1, 2026
2.1Average Edge Budget
9
Apr 1, 2026
0.78Average Edge Budget
9
Apr 1, 2026
1.72Average Edge Budget
9
Apr 1, 2026
0.78Average Edge Budget
9
Apr 1, 2026
60.95Accuracy
9
Mar 31, 2026
96Average Score
9
Mar 26, 2026
47F1 (Multi-Hop)
9
Mar 25, 2026
66.28Answer Accuracy
9
Mar 25, 2026
55.9ACCR
9
Mar 24, 2026
78.4ACCR (TQ)
9
Mar 24, 2026
80.2AUROC
9
Mar 24, 2026
86.8AUROC
9
Mar 24, 2026
0.87AUROC
9
Mar 24, 2026
70.6AUROC
9
Mar 24, 2026
97.5Hits@1
9
Mar 17, 2026
90.9Verification Accuracy (Exact-match)
9
Mar 17, 2026