Out-of-Distribution
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Out-of-Distribution OOD Avg (test)
78.4Accuracy (OOD Avg)
18
Out-of-Distribution Scatter Plot (test)
89.7Accuracy
18
Out-of-Distribution Bar Chart (test)
57Accuracy
18
Out-of-Distribution Large Table (test)
91.4Accuracy
18
Out-of-Distribution (ARC-c, GPQA, MMLU-Pro, LiveCodeBench)
96.6ARC-c (pass@1)
16
Out-of-Distribution (OOD) Split
29.31Mean Ratio
16
Out-of-Distribution Avg
59.7Avg Score (OOD)
15
Out of Distribution Overall
0.716SRCC
13
Out-of-distribution
96.2Return (r)
8
Out of Distribution (test)
78.4Harmony Ratio
5
Out-of-distribution GMM prior (test)
0.143Expected Distance (ED)
4
Out-of-Distribution (test)
92SG
2