PhD AI Research Intern (Contract) at Latitude in Remote

Summary: Latitude is a company focused on building AI-native games, starting with AI Dungeon. They are seeking a PhD AI Research Intern to conduct fundamental research on LLMs and create benchmarks for evaluating LLM behavior within their story engine. Responsibilities: - Conduct fundamental LLM research using our SOTA story engine - Create a benchmark for evaluating LLM behavior by defining tasks, success criteria, and auxiliary metrics within our story engine - Deliver a benchmark library for performing evaluations on SOTA LLMs and a written report of your compiled results for posting to our blog and potential publication Required Qualifications: - Current grad student in computer science or similar - Previous publications or research projects creating LLM benchmarks - Demonstrated software engineering experience via open-source projects or employment - Evidence of written communication skills via previous reports or publications Preferred Qualifications: - Actively pursuing PhD - First author publications at major conferences - Familiarity with Typescript Required Skills: Large Language Models (LLM), Benchmarking, Software Engineering, Typescript Benefits: Medical, dental, and vision coverage, with the company paying 100% of premiums for employees and eligible dependents (subject to plan terms and eligibility; benefits vary by location)., HSA options are available depending on the selected plan., Company-provided laptop (Mac by default) and the tools you need to do your best work., Flexible PTO, encouraged and supported., $100/month Learning Fund for books, courses, coaching, conferences, and video games.