Published several papers including 'How to Build Good Language Modeling Benchmarks', 'The Use Case for Relative Position Embeddings', etc.; shared experiences on finding research topics and advice for junior researchers.
Research Experience
Conducted eight years of research in the field of neural language models; involved in multiple projects such as building language modeling benchmarks, using relative position embeddings, etc.
Background
Research interests include neural language models and their improvement for human utility.
Miscellany
Believes that the most important parts of a PhD thesis are the acknowledgments and dedication sections; emphasizes the importance of honesty in research and maintaining a positive attitude towards failure.