Pushing the bounds of dropout

We show that dropout training is best understood as performing MAP estimation concurrently for a family of conditional models whose objectives are themselves lower bounded by the original dropout objective. This discovery allows us to pick any model from this family after training, which leads to a substantial improvement on regularisation-heavy language modelling. The family includes models that compute a power mean over the sampled dropout masks, and their less stochastic subvariants with tighter and higher lower bounds than the fully stochastic dropout objective. We argue that since the deterministic subvariant's bound is equal to its objective, and the highest amongst these models, the predominant view of it as a good approximation to MC averaging is misleading. Rather, deterministic dropout is the best available approximation to the true objective.

PDF Abstract ICLR 2019 PDF ICLR 2019 Abstract


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Language Modelling Penn Treebank (Word Level) 2-layer skip-LSTM + dropout tuning Validation perplexity 57.1 # 18
Test perplexity 55.3 # 23
Params 24M # 7