ResearchBenchmarkskNN Classification on TinyImageNet (val)Follow41.7kNN AccuracySimSiam35.77237.31138.8540.389Feb 13, 2025Evaluation ResultsMethodMethodLinkskNN AccuracySimSiamTraining Protocol=Cut...Training Protocol=Cut (c = 9), Epochs=5002025.0241.7SimSiamTraining Protocol=Defa...Training Protocol=Default, Epochs=5002025.0241.4SimCLRTraining Protocol=Grad...Training Protocol=GradScale, Epochs=5002025.0238.1SimCLRTraining Protocol=Cut...Training Protocol=Cut (c = 3), Epochs=5002025.0237.7SimCLRTraining Protocol=Defa...Training Protocol=Default, Epochs=5002025.0236