Our
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
our 132 molecules successfully solved (test)
7LRN
6
Our General scenarios (test)
1,108FVD
5
Our held-out Wikipedia tables (test)
0.79F1 Score
5
Our Eval. Set (300 prompts)
77.15Quality
2
Loading the SOTA2 catalog…