Rijul Jain
I’m an undergraduate at Williams College majoring in English and Computer Science. I was a visiting student at the University of Oxford for the 2023-24 academic year. I work on research at the intersection of programming languages, human-computer interaction, and artificial intelligence.
This summer, I designed and conducted a study of expert users of proof-oriented languages at Microsoft Research in Redmond, WA, under Shan Lu and Sarah Fakhoury. In the summer of 2023, I participated in Carnegie Mellon University’s Research Experiences for Undergraduates in Software Engineering (REUSE) program under Joshua Sunshine and Keenan Crane, researching domain-specific program generation with large language models to democratize diagram authoring using Penrose. As part of previous research with Daniel Barowy at Williams, I’ve also implemented a programming language that describes and captures UNIX filesystem state, sidestepping problems with breaking system call API changes to safely express program semantics for OS platform- and version-agnostic execution.
Over the course of the 2024-25 academic year, I’ll be working on two honors theses, one in computer science and one in English. With Joshua Sunshine and Bill Jannen, I will be developing testing strategies and tools to better characterize errors in and improve the usability of visualization DSLs. With Anita Sokolsky, I will be examining abrupt shifts in scale and proportion in the works of George Eliot, informed by the tensions between two aesthetic theories, those of Immanuel Kant and Theodor W. Adorno.
I’m broadly interested in ways for computer science research to learn from and interact with research attuned to the humanities. One manifestation of this, for example, focuses on the human-centered design of programming tools incorporating text-based generative AI interaction. Modeling programmers with research in cognitive science and psychology has proved fruitful; it may be that cognitive literary science, which, in part, uses cognitive psychology and neuroscience to explicate the mental processes and the corner cases of reading, will be relevant to modeling the interactions of end users with generative AI, given their often-text-based form. I hope to explore this and related topics further!
| resume | github | email me! |