Overall
Benchmarks
TaskDatasetSOTA resultTrendResults
Overall
89.6Accuracy
81
Overall
84.8Accuracy
40
Overall NQ, TriviaQA, BioASQ, PopQA
0.617Accuracy
32
Overall Combined Benchmarks
88.7Accuracy
31
Overall NaturalPlan AIME 2024 GPQA
96.5Final Score (Macro-Avg)
28
Overall 13 datasets aggregate
85.7N-Mean
26
Overall GSM8K, MATH-500, AMC, AIME24, AIME25
91.6Accuracy
26
Overall GSM8K, MATH-500, AMC 2023, AIME 2024, AIME 2025
79.8Accuracy
24
Overall Across five math reasoning datasets
45.8Overall Accuracy
24
Overall
93.51Accuracy
24
Overall MATH-500 AIME25 HumanEval GPQA
85.1Accuracy
24
Overall Combined 5 Datasets (test)
85.1mDice
24
Overall DB15K, MKG-W, MKG-Y
41.04MRR
22
Overall Aggregate
1.003Average Score
22
Overall Combined Benchmarks
49.64Performance (Seen Data)
21
Overall AMC23, AIME24, MATH500, GPQA-D aggregate
79.1Accuracy
21
Overall Combined Datasets
0.844mDice
21
Overall CSQA, StrategyQA, LogiQA
64.95Accuracy
20
Overall (Average)
80.5Recall@10
20
Overall Macro-average
70.97Accuracy (%)
20
Overall
62.05Overall Score
19
Overall
84.87Accuracy
19
Overall Dirichlet α=0.3 partition
70Accuracy
18
Overall 20 Chunks
61.22MAP
18
Overall 15 Chunks
59.99MAP
18