You can't pick your neighbors, or can you? When and how to rely on retrieval in the $k$NN-LM
Open paper
Benchmark
Model
Rank
Results
language-modelling-on-wikitext-103
kNN-LM w/ Adaptive Coefficient
#9
Test perplexity: 15.5
Validation perplexity: 15.72
…
Rank counts only results with a code link.