Luis A. Pabon

AA Graduate Student @Stanford

[Aiden Swann 🤠 CORL2026 @SwannAiden]

GPT-6 Astra's robotic capabilities are not emergent. It was trained on robotics specific data. So were GPT-6.1 Sol and Gemini Argon

Here's how we know 🧵

[Daniel Morton ✈️ CoRL2026 @danielpmorton]

Humanoid retargeting and IK in microseconds [no GPU required]

Orders of magnitude speedups, with perfect contact handling: accelerate your robots today at https://t.co/58MzJm0h9U

w/ @drmapavone

[Xilun Zhang ✈️ CoRL @XilunZhangXZ]

How can we build robotic systems that learn about their environment on their own?

🎱 Excited to share that our paper, “ Closing the Dynamics Gap through Self-Directed Robotic Interaction,” has been accepted to CoRL Late Breaking Results!

We started exploring how to close the dynamics gap in 2023 with COMPASS, followed by CAPTURE, focused on building accurate simulators for policy learning. With today’s more capable agents, we can let the agent develop the simulator and exploration process together. It decides not only how to update its model, but also which real-world interactions would help it understand the physics! 🤖🤖

#CORL2026 #Robotics

[Liam @ CoRL @Liam862373]

Can an agent learn a pool table’s physics by playing on it? 🎱

While agents can close the visual gap, the difference between real world and sim dynamics remains significant. For CoRL late breaking🤠, we let an agent take real shots to try and close this real2sim dynamics gap: https://t.co/g12jkuuB4n

Very fun short project with @XilunZhangXZ @MatthewFoutter @Luis_A_Pabon @drmapavone

[Daniel Morton ✈️ CoRL2026 @danielpmorton]

Having trouble keeping your humanoid robots under control? Luckily, our latest #CoRL2026 paper (w/ @drmapavone) keeps them in check: enforcing constraints never seen during training, while minimally interfering with the underlying RL policy

https://t.co/KSCVAwjkcF