Real or Fake Text?
Crowd-sourced gamification of evaluation for large-scale neural language models.
Predicting Orderliness Using Wikihow
Performing temporal event reasoning by fine-tuning BERT-based neural language models.
Leveraging cloud architectures to perform intelligence, surveillance, and reconnaissance. Published SPIE Defense 2016.
Evaluation Criteria for Human and Computer Written Text
Human-annotated taxonomy of both errors made by large-scale neural language models, and characteristics of human-written text.
GROVER: Generating Rap by Observing Verses
An LSTM-based with attention model written in PyTorch that uses the CMU pronounciation dictionary to generate rhyming lyrics with inflection and meter.