OpenAI Scholars 2019 Program Participants

OpenAI has announced its 2019 cohort of Scholars, bringing together a diverse group of researchers with backgrounds in physics, biology, statistics, and the humanities to advance AI research. This initiative integrates multidisciplinary expertise to tackle challenges in reinforcement learning (RL), natural language processing (NLP), and generative modeling.

Research Focus Areas and Scholar Contributions

Reinforcement Learning and Robotics

Several scholars are applying reinforcement learning to physical systems and complex decision-making processes:

  • Jonathan Michaux, a cell biologist and mathematician, is applying reinforcement learning to robot manipulation with the goal of designing algorithms for robotic locomotion and manipulation in real-world settings.
  • Elynn Chen, who holds a PhD in Statistics, focuses on deep RL and its applications within business management and healthcare.
  • Helen (Mengxin) Ji, a PhD student in Resource Economics and Statistics, is developing RL methodologies to be applied to sentiment analysis.
  • Yuhao Wan, with a background in Mathematics and Philosophy, is studying deep reinforcement learning with a specific interest in how learning methods exhibit generalization.

Natural Language Processing and Understanding

Research in the 2019 cohort emphasizes improving the reasoning capabilities and accessibility of language models:

  • Fatma Tarlaci, who holds a PhD in Comparative Literature and a Masters in Computer Science, is working on NLP methodologies to improve inference and reasoning in NLP.
  • Edgar Barraza, a physics graduate from Cornell University, is focusing his research on natural language understanding.

Generative Models and Multidisciplinary AI

The program includes researchers exploring the intersection of AI with visual data and human learning:

  • Janet Brown is investigating generative models and their capacity to identify critical features of data and images during the generation of reconstructions.
  • Nancy Otero, with a background in software engineering, math, psychology, and education, is developing AI prototypes intended to improve education and studying how AI redefines human learning.

Sources