My name is Robby Costales and I conduct research on artificial intelligence (AI), with the long-term aim of developing generally capable AI systems for the benefit of humanity and other sentient creatures. I am a computer science PhD student at the University of Southern California, advised by Prof. Stefanos Nikolaidis of the ICAROS lab, with previous guidance from Prof. Maja Matarić and Fei Sha.
Currently, my research focuses on enhancing reinforcement learning agents’ abilities to adapt and generalize over complex task distributions by introducing various structural inductive biases, with the aim of reducing the need for significant human supervision on which current AI methods heavily rely.
Previously, I was a student researcher at Google Brain where I was advised by Izzeddin Gür. Prior to joining USC, I studied at Columbia University where I was advised by Prof. Junfeng Yang, and also Bard College at Simon’s Rock, where I started my undergraduate.
- Sep 2022 - Multi-agent learning in composite tasks paper (ALMA) accepted at NeurIPS 2022
- May 2022 - In Mountain View working at Google Brain as a student researcher
- Feb 2022 - Paper on hierarchical affordances (HAL) accepted as spotlight at ICLR 2022
- Aug 2021 - I finally moved to Los Angeles!
- Aug 2020 - Started my PhD in computer science at the University of Southern California
- May 2020 - Graduated with bachelor’s in computer science from Columbia and Simon’s Rock
- May 2020 - Paper on live DNN trojan attacks presented at CVPR 2020 AdvML Workshop
- Jun 2019 - Working at Capital One over the summer as a data engineering intern
- Aug 2018 - Transferred to Columbia University‘s Fu Foundation School of Engineering
- Jun 2018 - At WUSTL working with Alvitta Ottley for CSE REU
- Jun 2017 - At University of Miami working with Vance Lemmon for CfS REU
- Dec 2016 - Learned about “for” loops in Intro to Computer Science :o
- Aug 2015 - Started college at Bard College at Simon’s Rock