COGS
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
COGS (generalization)
99Accuracy (Generalization)
25
COGS (test)
99Exact Match Accuracy
23
COGS
99.5Accuracy
22
COGS
95.84COGS EM
7
COGS
83.9Exact Match Accuracy
6
COGS nonce
81.4Accuracy
6