Krishnan Srinivasan

TRI's latest Large Behavior Model (LBM) paper landed on arxiv last night! Check out our project website: toyotaresearchinstitute.github.io/lbm1/ One of our main goals for this paper was to put out a very careful and thorough study on the topic to help people understand the state of the technology, and to share a lot of details for how we're achieving it. youtube.com/watch?v=BEXFnr…

Want your dexterous allegro hand to pick up a novel object, but don't know where to start? Check out our work "Get a Grip" at #corl2024, and grab our diverse grasping dataset and model for dexterous grasps :)

Introduce HumanPlus - Shadowing part Humanoids are born for using human data. We build a real-time shadowing system using a single RGB camera and a whole-body policy for cloning human motion. Examples: - boxing🥊 - playing the piano🎹/ping pong - tossing - typing Open-sourced!

We have a new ICML paper! Adaptive Horizon Actor Critic (AHAC). Joint work with @krishpopdesu @xujie7979 @eric_heiden @animesh_garg AHAC is a first-order model-based RL algorithm that learns high-dimensional tasks in minutes and outperforms PPO by 40%. 🧵(1/4)

We made a step-by-step video tutorial for building the UMI gripper! Please leave comments on @YouTube if you have any question youtu.be/x3ko0v_xwpg

# on technical accessibility One interesting observation I think back to often: - when I first published the micrograd repo, it got some traction on GitHub but then somewhat stagnated and it didn't seem that people cared much. - then I made the video building it from scratch, and the repo immediately went through hockey stick growth and became a verty often cited reference for people learning backpropagation. This was interesting because the micrograd code itself didn't change at all and it was up on GitHub for many months before, stagnating. The code made sense to me (because I wrote it), it was only ~200 lines of code, it was extensively commented in the .py files and in the Readme, so I thought surely it was clear and/or self-explanatory. I was very happy with myself about how minimal the code was for explaining backprop - it strips away a ton of complexity and just gets to the very heart of an autograd engine on one page of code. But others didn't seem to think so, so I just kind of brushed it off and moved on. Except it turned out that what stood in its way was "just" a matter of accessibility. When I made the video that built it and walked through it, it suddenly almost 100X'd the overall interest and engagement with that exact same piece of code. Not only from beginners in the field who needed the full intro and explanation, but even from more technical/expert friends, who I think could have understood it if they looked at it long enough, but were deterred by a barrier to entry. I think as technical people we have a strong bias to put up code or papers or the final thing and feel like things are mostly self-explanatory. It's there, and also it's commented, there is a Readme, so all is well, and if people don't engage then it's just because the thing is not good enough. But the reality is that there is still a large barrier to engage with your thing (even for other experts who might not feel like spending time/effort!), and you might be leaving somewhere 10-100X of the potential of that exact same piece of work on the table just because you haven't made it sufficiently accessible. TLDR: Step 1 build the thing. Step 2 build the ramp. 📈 Some voice in your head will tell you that this is not necessary, but it is wrong.

Foundational moment in robotics! Huge props to @QuanVng and @KarlPertsch for coordinating such a massive project

It's been great experience working with the team on this, with much more to come!

Check out the Robotics section (§2.3), discussing opportunities in applying #foundationmodels across the robotics pipeline. Challenges await! Collecting the right data, ensuring safety are crucial. But tackling these problems now – *before* building models – is key!

Making assistive teleoperation intuitive, easy to operate, and precise! The journal version of our work on the framework of learned latent actions + shared autonomy + personalization is out. We also have new studies with users with disability. Paper: arxiv.org/abs/2107.02907

Safe learning is critical for using RL to learn tasks in the real world. We introduce Recovery RL: arxiv.org/abs/2010.15920 in collaboration with @brthananjeyan, @SurajNair_1, Michael Luo, @krishpopdesu, @MinhoHwang6, @mejoeyg, @julianibarz, @chelseabfinn, @ken_goldberg 👇(1/8)

Today’s one of those days where if I suddenly woke up from a coma after >= 7mo and you explained everything that’s going on and why the sky turned completely orange, I’d demand to be put back into another coma

The latest issue Out-Of-Pocket from @nikillinit is so so on point. outofpocket.substack.com/p/the-college-…

I want naan, not NaN.

BREAKING: The Trump administration just agreed to rescind its harmful policy that would've barred international students who attend universities or colleges that are fully online this upcoming semester. This is what victory looks like. Keep speaking out. Keep fighting back.

Code now available at github.com/TeaganLi/ICRA2…

@NAChristakis @threadreaderapp unroll

Mass gatherings facilitate transmission in this study which investigates the seroprevalence in Gangelt, Germany, among those attending carnival celebrations, finding a 2.5X increase in rate of infection (they were not masked, however). medrxiv.org/content/10.110… 32/

I want to go on record with obvious point: large gatherings of people facilitate spread of contagious disease. The *reason* for the gathering (whether street protests for a cause I support, or GOP convention, or a sporting event) is not material to the spread of the virus. 1/

The International Conference on Robotics and Automation (ICRA) 2020 is being hosted virtually from May 31 – Jun 4. We’re excited to share all the work from SAIL and Stanford that’s being presented in our new blog post: ai.stanford.edu/blog/icra-2020/

Here are the problem-solving tips I went through yesterday. Each may seem simple, but it’s shocking how often they help to get unstuck. What would you add?

Kudos to Tingguang Li, @krishpopdesu, Max Qing-Hu Meng and Wenzhen Yuan. ArXiv arxiv.org/abs/1910.10985 Project sites.google.com/view/learningh… Video youtu.be/s8j2b79ByuQ