Tom Silver

Assistant Professor @Princeton. Developing robots that plan and learn to help people.

[Tom Silver @tomssilver]

We need better terminology to distinguish these!

[Tom Silver @tomssilver]

At #CoRL2026 I'm hearing several variations on "agents may be effective at robot decision-making, but they’re too slow" or "not reactive".

[E.g., motivating VLAs/WAMs.]

I think this means at least 3 different things depending on how the agent is being used. 🧵

[Tom Silver @tomssilver]

Similar confusion for "agents aren’t closed-loop / reactive." True in (a). But in (b) and (c) the agent outputs a program, and a program can be a closed-loop policy running at control rates.

[Tom Silver @tomssilver]

I'll be around the #CoRL2026 Late Breaking session in 20 minutes with @Bw_Li1024, @WangRobin58334, and @YixuanHuang13 to chat about this!

[Matteo Merler @merler_m]

✈️If you are at @corl_conf today, make sure to stop by the Late Breaking Results posters at 1 pm EST, where my amazing co-authors will be presenting our work!

🔗 Paper, videos and new Opus 5.5 results: https://t.co/znk0mOlJjH

[Tom Silver @tomssilver]

This week's #PaperILike is "Principles and Applications of Continual Computation" (Eric Horvitz [yes, that one!], 2001).

An early work on "sleep time" compute, increasingly relevant for agents (e.g., https://t.co/33xbTaOU8j).

PDF: https://t.co/X3LFy2WMgY

[Tom Silver @tomssilver]

A Case for Paper Grading Boards (Or: A Plea for More Proposals to Radically Reform Peer Review)

https://t.co/OL2whHHFOu

[Tom Silver @tomssilver]

This week's #PaperILike is "On-line Policy Improvement using Monte-Carlo Search" (Tesauro & Galperin, 1996).

A classic example of combined learning and planning, a precursor to AlphaZero (in backgammon), and an easy read.

PDF: https://t.co/MRGGLJw8n6

[Kartik Garg @_pacificax]

Sprout Jackson

[Robin Wang @WangRobin58334]

Excited to share UniPred, accepted to the IEEE T-RO 2026(SI on FM for Robotics)! Robots plan long tasks with concepts, but demonstrations only show actions. UniPred learn how to invents those concepts with agent in the loop. 🧵
More in: https://t.co/3RaO7hTeAA

[Abhay Deshpande @ CoRL 2026 @ab_deshpande]

DataFarm is cool! Making procedurally generated data look less "robotic" makes a lot of sense. Here's a lightweight JAX/PyRoki port if people are interested, should make it easier to use outside of the TiPToP framework

https://t.co/6LoYeMckrH

[Tom Silver @tomssilver]

This week's #PaperILike is "Human-robot collaborative task planning using anticipatory brain responses" (Ehrlich et al., PLOS One 2023).

Talking is too much effort. I want robots to read my mind! (+nice HRI / collaborative planning insights)

PDF: https://t.co/rEYoCL9tSS

[Toru @ToruO_O]

How should we approach research, when the very field we work on is going through rapid and potentially fundamental changes?

I recently gave a talk at the Stanford Vision Seminar, reflecting on my PhD journey and how AI is changing the way I think about research. I adapted part of it into a new blog post, organized around three questions:
1. What should we work on?
2. How should we work with AI?
3. What is a PhD for?

I hope some of my reflections are helpful. The talk ended with an open discussion, and I’d love to continue it here; curious how others are thinking about these questions.

Personally -- as much as there's FOMO, I also think life is very long, and we have time.

https://t.co/DytwGOzBLy

[Tom Silver @tomssilver]

@pranavnnt ¿Por qué no los dos?

[Tom Silver @tomssilver]

Adding this to my catalogue of (very) surprising physical strategies discovered by coding agents...

[Bowen Li @Bw_Li1024]

Me: Put the green box on target, clearing obstructions.
Astra: You got it, boss 🫡

(Purely python solution generated in a red teamed sandbox using no any other tools, with $20 budget)

[Tom Silver @tomssilver]

This week's #PaperILike is a break from the usual: "Worms and science," a 2003 interview with Sydney Brenner.

On choosing problems, hypothesis-driven vs data-driven science, and knowing when to move on (in his case, from molecular bio).

PDF: https://t.co/WEm5PszlCM

[samrat @samratsahoo_]

Teleoperators collecting data to fine-tune VLAs are spending valuable time demonstrating actions robots already know how to do.

TANDEM combines task and motion planning (TAMP) with human teleoperation only as needed, reducing human effort.

(1/8)

[Yixuan Huang @YixuanHuang13]

😫Tired of spending hours teleoperating robots?

🚀What if you could collect 2.9× more demonstrations with the same human effort?

Introducing TANDEM: our new work that combines autonomous task and motion planning (TAMP) with human teleoperation only when needed. 🧵 1/n

[Tom Silver @tomssilver]

Check out Samrat’s extension to DATAFARM where there is now a human in the loop to do what TAMP alone cannot!

[samrat @samratsahoo_]

Teleoperators collecting data to fine-tune VLAs are spending valuable time demonstrating actions robots already know how to do.

TANDEM combines task and motion planning (TAMP) with human teleoperation only as needed, reducing human effort.

(1/8)