Tucker Hermans
@thermans.bsky.social
Professor of Computing, I do research on robotics and AI. I also ski.
https://robot-learning.cs.utah.edu/thermans
https://robot-learning.cs.utah.edu/thermans
Pinned
Tucker Hermans
@thermans.bsky.social
· May 29
Probabilistic Deep Learning - Public Overview - Tucker Hermans - Kahlert School of Computing; University of Utah
Probabilistic Deep Learning This class covers probabilistic approaches to deep learning. These include model calibration, mixture density networks, (approximate) Bayesian neural networks, variational ...
docs.google.com
A few folks at ICRA last week were asking about the details of my probabilistic deep learning course so I through together a public syllabus for it: docs.google.com/document/d/1...
Reposted by Tucker Hermans
Come be my colleague! utah.peopleadmin.com/postings/188... . Our school (we're bigger than a department) is an awesome place to work. Our faculty are excellent across the board. Pay and benefits are good. SLC is a nice place to live. Happy to answer any questions
Assistant Professor (Lecturer) in Computing
utah.peopleadmin.com
October 14, 2025 at 7:36 PM
Come be my colleague! utah.peopleadmin.com/postings/188... . Our school (we're bigger than a department) is an awesome place to work. Our faculty are excellent across the board. Pay and benefits are good. SLC is a nice place to live. Happy to answer any questions
Our cuTAMP code is now available: github.com/NVlabs/cuTAMP
GitHub - NVlabs/cuTAMP: cuTAMP: Differentiable GPU-Parallelized Task and Motion Planning
cuTAMP: Differentiable GPU-Parallelized Task and Motion Planning - NVlabs/cuTAMP
github.com
August 5, 2025 at 6:08 PM
Our cuTAMP code is now available: github.com/NVlabs/cuTAMP
Hope to see many friends here! We'll be presenting cuTAMP: cutamp.github.io
June 14, 2025 at 3:21 AM
Hope to see many friends here! We'll be presenting cuTAMP: cutamp.github.io
A few folks at ICRA last week were asking about the details of my probabilistic deep learning course so I through together a public syllabus for it: docs.google.com/document/d/1...
Probabilistic Deep Learning - Public Overview - Tucker Hermans - Kahlert School of Computing; University of Utah
Probabilistic Deep Learning This class covers probabilistic approaches to deep learning. These include model calibration, mixture density networks, (approximate) Bayesian neural networks, variational ...
docs.google.com
May 29, 2025 at 7:03 PM
A few folks at ICRA last week were asking about the details of my probabilistic deep learning course so I through together a public syllabus for it: docs.google.com/document/d/1...
Got an email from NSF that my REU supplement won't be funded but I first read it as my entire RI Large was cancelled. 😅
May 23, 2025 at 4:43 PM
Got an email from NSF that my REU supplement won't be funded but I first read it as my entire RI Large was cancelled. 😅
This is great and the kind of close reading analysis I think a VLM is far from being able to produce anytime soon.
This is a little bit genius: an analysis of OpenAI based on a close reading of a 22-second video of CEO Sam Altman cooking pasta. “His kitchen is a catalogue of inefficiency, incomprehension, and waste.” [ft.com]
Three things we learned about Sam Altman by scoping his kitchen
All Drizzle, no sizzle
www.ft.com
May 13, 2025 at 12:02 AM
This is great and the kind of close reading analysis I think a VLM is far from being able to produce anytime soon.
I just watched Arrival (one of my favorites) on my flight back from the NVIDIA Research Onsite and I feel inspired to create a heptapod VLA.
May 8, 2025 at 2:23 AM
I just watched Arrival (one of my favorites) on my flight back from the NVIDIA Research Onsite and I feel inspired to create a heptapod VLA.
There's nothing like running an MCMC algorithm to remind you that computers can do a lot more than just run web browsers and PowerPoint.
May 5, 2025 at 7:01 PM
There's nothing like running an MCMC algorithm to remind you that computers can do a lot more than just run web browsers and PowerPoint.
Currently writing an active learning (pedagogy) exercise about active learning (ML concept)
April 1, 2025 at 5:59 PM
Currently writing an active learning (pedagogy) exercise about active learning (ML concept)
Saw a flyer for this on campus today! @standupforscience.bsky.social
March 4, 2025 at 2:50 AM
Saw a flyer for this on campus today! @standupforscience.bsky.social
Just got word from a colleague that he's been fired from NSF. It sounds like all at-will and probationary employees were let go, repeating what happened at NIH last week.
February 18, 2025 at 4:58 PM
Just got word from a colleague that he's been fired from NSF. It sounds like all at-will and probationary employees were let go, repeating what happened at NIH last week.
Norman's work influenced my thinking about affordnaces for robot learning in my dissertation. I'm excited for this!
Mark your calendars for this Feb. 27 discussion with Don Norman, a renowned expert in user-centered design and beyond. He's among the most cited researchers in the world. 🔥
Norman will be hosted by Vineet Pandey, one of the inaugural faculty fellows in SCI's One-U Responsible AI Initiative.
Norman will be hosted by Vineet Pandey, one of the inaugural faculty fellows in SCI's One-U Responsible AI Initiative.
February 15, 2025 at 1:02 AM
Norman's work influenced my thinking about affordnaces for robot learning in my dissertation. I'm excited for this!
So nice to see an organization I’m a member of, AAUP, actually doing something to fight back against the attack on science! ✊
BREAKING: New lawsuit challenges Trump's anti-diversity executive orders, alleging Spending Clause, vagueness (Due Process), free speech (First Amendment), and separation of powers violations. storage.courtlistener.com/recap/gov.us...
February 4, 2025 at 2:55 AM
So nice to see an organization I’m a member of, AAUP, actually doing something to fight back against the attack on science! ✊
Reposted by Tucker Hermans
NSF is still freezing grant payments because of Trump's executive orders targeting DEI, despite NSF's congressional mandate to weigh how grants will expand participation in science.
What that means for scientists/science (with more detail than yesterday's radio story)
www.npr.org/2025/01/31/n...
What that means for scientists/science (with more detail than yesterday's radio story)
www.npr.org/2025/01/31/n...
National Science Foundation freezes payments in response to Trump's executive actions
The National Science Foundation's funding freeze, and wider confusion about the status and future of science funding, is already hampering research.
www.npr.org
January 31, 2025 at 5:53 PM
NSF is still freezing grant payments because of Trump's executive orders targeting DEI, despite NSF's congressional mandate to weigh how grants will expand participation in science.
What that means for scientists/science (with more detail than yesterday's radio story)
www.npr.org/2025/01/31/n...
What that means for scientists/science (with more detail than yesterday's radio story)
www.npr.org/2025/01/31/n...
I just now got an official email from NSF about the freezing of new and existing grants, the relevant highlight:
"Therefore, all review panels, new awards, and all payments of funds under open awards will be paused as the agency conducts the required reviews and analysis."
"Therefore, all review panels, new awards, and all payments of funds under open awards will be paused as the agency conducts the required reviews and analysis."
January 29, 2025 at 4:41 AM
I just now got an official email from NSF about the freezing of new and existing grants, the relevant highlight:
"Therefore, all review panels, new awards, and all payments of funds under open awards will be paused as the agency conducts the required reviews and analysis."
"Therefore, all review panels, new awards, and all payments of funds under open awards will be paused as the agency conducts the required reviews and analysis."
Our paper Points2Plans was accepted to ICRA. Dope. sites.google.com/stanford.edu...
January 27, 2025 at 10:11 PM
Our paper Points2Plans was accepted to ICRA. Dope. sites.google.com/stanford.edu...
Reposted by Tucker Hermans
If you're a scientist that serves on an NSF panel that got cancelled today, I'd love to talk to you about it for an NPR story. DM, email or reach out on signal.
January 27, 2025 at 5:37 PM
If you're a scientist that serves on an NSF panel that got cancelled today, I'd love to talk to you about it for an NPR story. DM, email or reach out on signal.
When I posted about my class I thought I'd give a thought about class each day, that's too much for me right now. Instead I'll say we are covering ensembles this week and it's been fun!
January 23, 2025 at 5:06 AM
When I posted about my class I thought I'd give a thought about class each day, that's too much for me right now. Instead I'll say we are covering ensembles this week and it's been fun!
I'm teaching a new class this semester called "Probabilistic Deep Learning." I've organized the topics roughly around different probabilistic neural network tools we've used in my group. I'm very excited to be teaching a class that doesn't require students to install simulators or use robots!
January 9, 2025 at 5:35 PM
I'm teaching a new class this semester called "Probabilistic Deep Learning." I've organized the topics roughly around different probabilistic neural network tools we've used in my group. I'm very excited to be teaching a class that doesn't require students to install simulators or use robots!