Zum Inhalt springen
Global sound
RoboPapers

RoboPapers

Ep#103: Freeform Preference Learning for Robotic Manipulation

· 48 Min.

Shownotes

Specifying reward functions for robots is one of the hardest things about reinforcement learning. Robot rewards often need to be very detailed; metrics like progress can be ill-defined and hard to estimate. This leads to most robot learning defaulting to sparse rewards or simple preference learning. But naive preference learning (having human annotators choose one trajectory over another) is an easy solution, but obscures a lot of the signal in complex tasks and can make learning a lot less efficient.
Marcel Torné, Anubha Mahajan, and Abhijnya Bhat join us to talk about their solution…

Website: robopapers.substack.com

Live, letzte Stunde: 15.573 Songs auf 7.821 Sendern in 148 Ländern

Entdecke die Welt über das Radio

Mehr zum Entdecken (9)

Etwas Schönes gehört? Speicher es oder öffne es direkt in Spotify.

★ Artist of the week The fastest riser on the radio this week
Your song here? From € 9 per week →
Rus Radio LithuaniaWeiterhören