Considering RNNs have completely taken over text generation with language models, are there any papers that comprehensively compare the power of RNN and Markov models?
This is not a paper, but an interesting blog post nevertheless:
The unreasonable effectiveness of Character-level Language Models (and why RNNs are still cool)
https://nbviewer.jupyter.org/gist/yoavg/d76121dfde2618422139