Blake Wulfe retweeted
How can you know when you don’t know? At Walden, our robots continuously learn more than just tasks from deployments: they also learn when they should ask for help.
This requires a fundamentally different full-stack approach to data, training, inference, and operations. The payoff is accelerated deployments at the frontier, showing compounding in the field already today. And customers going from “how long will it take” to “can it do this next?”
See how: waldenrobotics.com/news/call…
Blake Wulfe retweeted
I’m super excited to share one of the new capabilities we’ve been working on at @WaldenRobotics
Our robots learn and improve every day on the job. When they encounter something new or aren’t sure how to proceed, they call on a remote assistant who steps in to help.
This lets us deploy at the edge of our robots’ capabilities, collect the most useful data in the field, and scale our fleet faster.
More details soon!
When in doubt, phone a friend ☎️
At @WaldenRobotics, our policies know when to ask for help.
Here, the robot tries scanning the wrong side of a transparent product and calls a remote assistant. The intervention gets it unstuck and becomes training data for the next deployment.
Blake Wulfe retweeted
When in doubt, phone a friend ☎️
At @WaldenRobotics, our policies know when to ask for help.
Here, the robot tries scanning the wrong side of a transparent product and calls a remote assistant. The intervention gets it unstuck and becomes training data for the next deployment.
Blake Wulfe retweeted
In January, I started "building something new" with an incredible team. Today I finally get to share some first details about what we've been building. We've called it Walden Robotics (waldenrobotics.com).
I thought long and hard about my own reasons for starting this company. It's not only about the robots. It's also about people. I've tried to capture those thoughts in my first Walden blog post: waldenrobotics.com/news/why-…
It's been an incredible ride so far. Within just a few months of forming the company, we were already operating a general-purpose robot with an end-to-end policy in production in one of the most important factories in North America. It's amazing at how much I've already learned from that experience. There is a lot of work to do, but the mission has never been so clear.
Please help me welcome Walden Robotics into the world. And stay tuned for more updates!
youtube.com/watch?v=fewvZrck…
Blake Wulfe retweeted
Announcing OpenThinker3-7B, the new SOTA open-data 7B reasoning model: improving over DeepSeek-R1-Distill-Qwen-7B by 33% on average over code, science, and math evals.
We also release our dataset, OpenThoughts3-1.2M, which is the best open reasoning dataset across all data scales. Full details are in our ✨new paper✨ - below we share the highlights:
BTW, it also works on non-Qwen models😉 (1/N)
Blake Wulfe retweeted
Meet ProVox: a proactive robot teammate that gets you 🤖❤️🔥
ProVox models your goals and expectations before a task starts — enabling personalized, proactive help for smoother, more natural collaboration. All powered by LLM commonsense.
Recently accepted at @ieeeras R-AL!
🧵1/7
Blake Wulfe retweeted
Image/video models can transfer knowledge from Internet data to robot agents by generating goal images. But what happens when images have harmful visual artifacts? We present GHIL-Glue, a method to align image/video models and low-level policies.
ghil-glue.github.io/
Blake Wulfe retweeted
Learned visuomotor robot policies are sensitive to observation viewpoint shifts, which happen all the time. Can visual priors from large-scale data help? Introducing VISTA: using zero-shot novel view synthesis models for view-robust policy learning! #CoRL2024
🧵👇
Blake Wulfe retweeted
After two years, it is my pleasure to introduce “DROID: A Large-Scale In-the-Wild Robot Manipulation Dataset”
DROID is the most diverse robotic interaction dataset ever released, including 385 hours of data collected across 564 diverse scenes in real-world households and offices
Blake Wulfe retweeted
Access to *diverse* training data is a major bottleneck in robot learning. We're releasing DROID, a large-scale in-the-wild manipulation dataset. 76k trajectories, 500+ scenes, multi-view stereo, language annotations etc
Check it out & download today!
💻: droid-dataset.github.io/
Blake Wulfe retweeted
ICRA 2022 AV workshop videos now online:
🚗 keynotes: icra2022av.org/#h.198firfh7f…
🚗 authors: icra2022av.org/#h.r8uf7ehwpn…
🚗 all: youtu.be/qzo61V7G1EM?t=538
Thanks to lead organizer @corina_gurau, virtual org @wulfebw @hu_anth @pondruska @blazejosinski, PC members + authors + attendees!
Blake Wulfe retweeted
How should models be trained? We show weighting model errors by their effect on downstream control improves accuracy where it matters most (gif) + improves end performance
arxiv.org/abs/2204.13319
youtu.be/OSzh6Jkp5io
w @wulfebw @MercatJean @logan_m_ellis @svlevine @adnothing
Blake Wulfe retweeted
*Deadline Extension* Call for a second round of submissions to the #ICRA2022 workshop on Fresh Perspectives on the Future of Autonomous Driving.
Submit by April 25 AoE: icra2022av.org/
@rowantmc @hu_anth @wulfebw @pondruska @felipealcm @nick_rhinehart @szagoruyko5
Blake Wulfe retweeted
Calling papers for the @ieee_ras_icra workshop on Fresh Perspectives on the Future of Autonomous Driving in May!
Submit by 22 March: self-driving-cars.org/worksh…
with @corina_gurau @hu_anth @szagoruyko5 @felipealcm @nick_rhinehart @pondruska @wulfebw Błażej Osiński #ICRA2022