Download PDF
Conference Articles
K.R. Zentner , Ryan C. Julian, Brian Ichter, Gaurav S. Sukhatme,
Conditionally combining robot skills using large language models
,
2023, (ICRA 2024) [URL]
K.R. Zentner , Ujjwal Puri, Yulun Zhang, Ryan Julian, Gaurav S. Sukhatme,
Efficient multi-task learning via iterated single-task transfer
,
2022, (Oral Presentation at IROS 2022) [URL]
K.R. Zentner , Ujjwal Puri, Zhehui Huang, Gaurav S. Sukhatme,
Guaranteed trust region optimization via two-phase KL penalization
,
2023, (Under Preparation) [URL]
Shashank Hegde, Sumeet Batra, K.R. Zentner , Gaurav S. Sukhatme,
Generating behaviorally diverse policies with latent diffusion models
,
2023, (Poster at NeurIPS 2023) [URL]
Other Research
K.R. Zentner ,
doexp: A simple experiment runner
,
2023, [URL]
K.R. Zentner ,
Language-World: A language and robotics benchmark based on Meta-World
,
2023, [URL]
K.R. Zentner , Ryan Julian, Ujjwal Puri, Yulun Zhang, Gaurav S. Sukhatme,
Towards exploiting geometry and time for fast off-distribution adaptation in multi-task robot learning
,
2021, (Poster at the NeurIPS 2020 Workshop: Challenges of Real World Reinforcement Learning and NeurIPS 2020 Workshop: Offline RL) [URL]
K.R. Zentner , Ujjwal Puri, Yulun Zhang, Ryan Julian, Gaurav S. Sukhatme,
Multi-task learning via iterated single-task transfer
,
2022, (Poster Presentation at The Multi-disciplinary Conference on Reinforcement Learning and Decision Making 2022) [URL]
K.R. Zentner , Ryan Julian, Ujjwal Puri, Yulun Zhang, Gaurav S. Sukhatme,
A simple approach to continual learning by transferring skill parameters
,
2021, [URL]
The garage contributors,
Garage: A toolkit for reproducible reinforcement learning research
,
2019, (Contributor 2018-2023) [URL]
Tianhe Yu, Deirdre Quillen, Zhanpeng He, Ryan Julian, Avnish Narayan, Hayden Shively, Adithya Bellathur, Karol Hausman, Chelsea Finn, Sergey Levine,
MetaWorld
,
2019, (Lead Maintainer 2021-2023) [URL]
Career History
In short: I attended UC Berkeley (BS EECS, 2015), worked for Google (X, Search, Brain), and now attend USC (PhD CS, 2024).
University of Southern California
I’m currently a PhD student in computer science, advised by Gaurav Sukhatme. I’ve published several papers, which you can see here. I’ve worked as a Research Assistant for much of my time at USC in the Robotic Embedded Systems Laboratory.
Google Brain (now Google DeepMind)
I interned at Google Brain in Summer 2022, which eventually lead to a paper on controlling robots via a fine-tuned interpretation of plans generated by a Large Language Model.
Google Search
I worked on the Google Search Infrastructure team from 2016 to 2018. My largest accomplishment was a client-side caching layer which reduced bandwidth used for JavaScript by 50% on Tier 1 searches, while measurably reducing latency and reducing global power usage by roughly 10MW.
Google Robotics / X (now Everyday Robots)
I interned with this group in Spring 2016, when the Everyday Robots project was first started. As my intern project, I created the first web-based data visualizer (“v0”). I designed the overall architecture and implemented video and compressed voxel visualization.
University of California, Berkeley
I received a Bachelors of Science in Electrical Engineering and Computer Science from UC Berkeley. While I was there, I worked as a TA for The Beauty and Joy of Computing, lead the GamesCrafters research group, and was heavily involved in PiE Robotics.
PiE Robotics
I took on several volunteer roles at PiE, participating in over a hundred work session and helping to host dozens of on-campus events.
My first year with them, I helped develop their in-house open-hardware motor controller, the Grizzly Bear. This allowed students to use several more motors on their robots. I also developed their field control system, Forsetti.
My second year with PiE I helped lead Control System Next, an ambitious project to build an extremely inexpensive yet modular real-time robotic control system.
My third year with PiE I served as Engineering Coordinator, where I managed roughly 40 volunteer staff for a semester. I also taught at a high school in Oakland once a week as a PREP mentor.
Mozilla
I interned at Mozilla in the summer of 2015. I implemented an early version of the CSS Containment specification, and recommended ways for the specification to be improved. I gained in-depth knowledge about HTML layout, which lead me to contribute the original flexbox parser to Servo.
I also fixed a subtle bug in HTML table layout that had been present since HTML tables were first implemented.
Wind River / VxWorks CoreOS
I interned with the VxWorks CoreOS team in the summer of 2013. I added directory support to the NetDRV file system, fixed a few bugs, and improved the POSIX compatibility layer.
Qualcomm
I intern here in the summer of 2014. I implemented a C parser and binding generator for running TD-CDMA protocol tests, a GUI to a set of version control scripts, and a test automation server.