yvesborensteininunpackPerplexity and accuracy in classificationPerplexity is a metric used essentially for language models. But since it is defined as the exponential of the model’s cross entropy, why…Feb 1, 2021Feb 1, 2021
yvesborensteininunpackReflection on EmbeddingsHere are some brief looks and suggestions of further explorations as to the information that embeddings of categorical data can yield…Jan 18, 2021Jan 18, 2021
