Average
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Average (PatternNet, RSICD, RESISC45, MLRSNet)
92.52Base Score
8
Average
65.08Precision
8
Average Overall
2.49Speedup
8
Average (Rain200L, Rain200H, DID, DDN, SPA)
38.33PSNR
8
Average cross-dataset (unseen)
29.3Success Rate (SR)
8
Average PlaceToast, TurnOnFaucet, MakeCoffee
90.7TSR
7
Average Task Order 3
92.42Accuracy (CA)
7
Average Task Order 2
85.39Clean Accuracy
7
Average (SST2, AGNEWS, GSM8K, AlpacaEval)
35.5HS Score
7
Average over 10 datasets
79.03Base Score
7
Average (5 datasets) Macro-average (test)
89.3Base Accuracy
7
Average (GSM8K-CoT, MATH, MBPP, HumanEval)
51.6Accuracy
7
Average Overall
32.7EM (Exact Match)
7
Average
99ASR
7
Average (HellaSwag, PiQA, OBQA, COPA, LogiQA, WinoG, SciQ, ARC-E, Lambada)
42.5Accuracy
7
Average All Datasets
86.05DSC
7
Average
66.2Accuracy
7
Average
29.07PSNR
7
Average (SHIQ, PSD, NSH)
36.059PSNR
7
Average webw., hle, gaia
9.87Accuracy
7
Average Across all 11 ultrasound sites (test)
74.39mIoU
7
Average all shapes
99.63IoU
6
Average (mMARCO-zh, HotpotQA, NQ, TQA)
66.55NDCG@10
6
Average Overall
71.3P@4
6
Average 4 Math Benchmarks
56.46Pass@1 Score
6