Self-Normalization Properties of Language Modeling
This addresses computational efficiency in language modeling for NLP applications, but is incremental as it builds on existing methods like NCE and softmax.
The study investigated self-normalization in language models to avoid computing partition functions, finding a surprising negative correlation between self-normalization and perplexity across models.
Self-normalizing discriminative models approximate the normalized probability of a class without having to compute the partition function. In the context of language modeling, this property is particularly appealing as it may significantly reduce run-times due to large word vocabularies. In this study, we provide a comprehensive investigation of language modeling self-normalization. First, we theoretically analyze the inherent self-normalization properties of Noise Contrastive Estimation (NCE) language models. Then, we compare them empirically to softmax-based approaches, which are self-normalized using explicit regularization, and suggest a hybrid model with compelling properties. Finally, we uncover a surprising negative correlation between self-normalization and perplexity across the board, as well as some regularity in the observed errors, which may potentially be used for improving self-normalization algorithms in the future.