Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

NevIR

Benchmarks

Task NameDataset NameSOTA ResultTrend
Information Retrieval ReasoningNevIR
Pairwise Accuracy0.617
14
Question AnsweringNevIR 20 randomly sampled queries
Accuracy40
8
Information RetrievalNevIR
V@243
8
Negation RetrievalNevIR
Pairwise Score29.28
6
Showing 4 of 4 rows