Tjeerd Boonstra
@tjeerdwboonstra.bsky.social
Neuroscientist at Maastricht University. Complex networks, brain-body interactions and motor control. Editor for Imaging Neuroscience
Poeh… Henk Kamp doet de VVD vanavond geen plezier (en ook geen positieve bijdrage aan Pauw & De Wit). Echt moeilijk om aan te horen. Wie dacht dat dit een goed idee was?
November 11, 2025 at 10:00 PM
Poeh… Henk Kamp doet de VVD vanavond geen plezier (en ook geen positieve bijdrage aan Pauw & De Wit). Echt moeilijk om aan te horen. Wie dacht dat dit een goed idee was?
Reposted by Tjeerd Boonstra
We developed Arnold - the first control policy that learns multiple motor skills across detailed musculoskeletal models. A small step towards understanding the computational principles behind motor intelligence.
Arnold: a generalist muscle transformer policy
Controlling high-dimensional and nonlinear musculoskeletal models of the human body is a foundational scientific challenge. Recent machine learning breakthroughs have heralded policies that master ind...
www.arxiv.org
September 4, 2025 at 8:45 AM
We developed Arnold - the first control policy that learns multiple motor skills across detailed musculoskeletal models. A small step towards understanding the computational principles behind motor intelligence.
Congrats on the nice paper! Haven't read it yet. Do you mean predict in a correlational or causal manner?
July 30, 2025 at 12:18 PM
Congrats on the nice paper! Haven't read it yet. Do you mean predict in a correlational or causal manner?
My feeling is that it is stimulated by shallow contact with many students and countered by in-depth contact with a few. If you’re coordinating entire cohorts you mostly hear about the negative outliers
July 19, 2025 at 8:50 AM
My feeling is that it is stimulated by shallow contact with many students and countered by in-depth contact with a few. If you’re coordinating entire cohorts you mostly hear about the negative outliers