Sarvesh Patil

This account is verified. Learn more

@servo97

Your friendly neighborhood PhD in Robotics @CMU.
Dexterous Manipulation | Generative Control | Reinforcement Learning

Pittsburgh, PA servo97.github.io Joined November 2014

652 Following 669 Followers 917 Posts

Latest posts 5

Pinned post

Sarvesh Patil @servo97 Jun 30

Interaction with the real world is the major bottleneck in robot learning. So what would robot RL look like if we didn’t need to limit compute per interaction? Our latest work, Off-Policy Generative Policy Optimization (OGPO, accepted to ICML26) embarks on answering this question (spoiler alert: when done correctly, it helps massively!).
🧵(1/N)

Sarvesh Patil @servo97 2d

Very fun collaboration on understanding how semantic generalizability is observed in multi-task VLA-RL for manipulation by scaling Q-function training.

RL post-training is often viewed as a sharpening mechanism that overfits to training task distributions and fails to generalize in OOD settings. But what if that's just a scale issue?

Atleast for semantic variations in simple manipulation tasks this seems to be case. Multi-task training improves performance on unseen scenes and objects with no policy updates (Best of N sampling via Q-values) and even more with policy extraction updates (OGPO, AWR)!

Michal Korniak @MichalAKorniak 3d

Can RL post-training of generalist robot policies enable improvement not just on the tasks we train on, but also generalization beyond them, improving on novel tasks not seen in training?

We show that, with the proper RL recipe, we can enable such "generalist improvement"! 🧵

View quoted post on X

Sarvesh Patil @servo97 Oct 20

Probably the cutest home robot I've seen so far! 🥹

Ros Shinkle Verified account @make_robot_go Oct 20

Introducing Voxel, a droid butler made by Tuesday Labs.

View quoted post on X

Sarvesh Patil @servo97 Oct 14

Video model encodings with recurrent architectures (e.g. GDNs) for latent memory compression (like workspace models) seem to be scaling with the use of world models at scale in multi-task settings!

The most impressive tasks for me are the seemingly innocuous ones, like wiping a table. Purely inferring actions from states that look more or less the same can render traditional MDP formulations unsuccessful. Using latent memory states to steer policy actions seems neat to convert the POMDP to an MDP!

Yilin Wu ✈️ CoRL 2026 Verified account @YilinWu11 Oct 14

What to remember in the past to act—spatial, temporal or motion details?

We often assume what’s useful, but that’s hard to know in advance.

Meet IMPRINT 🧠: retain rich memory first, then learn to use it for control. Compact latents keep inference fast as history grows. 🧵 1/10

View quoted post on X

Sarvesh Patil @servo97 Oct 1

Very cool work on distilling deliberative supervision from powerful but slow reasoning models into a "latent agentic harness" during training time!

These embedding spaces can then be used at inference time for policy steering with minimal latency, e.g. as latent memory modules for learned robot policies to recall and/or react to environment!

Nitish Dashora @ CoRL Verified account @DashoraNitish Oct 1

There’s lots of buzz around agentic harnesses for robots, particularly for long horizon tasks that require complex reasoning and memory. But what will it really take to turn a reasoning agent into a reactive, reliable and low-latency robot policy?

In our new paper, workspace models, we design a new memory architecture that acts as a latent harness for stronger reasoning models. Here’s why we think this might be the way forward:

(1/9)

View quoted post on X