Siirry sisältöön
Global sound
RoboPapers

RoboPapers

Ep#105: Beyond Imitation: Self-Improving Robot Policies via Off-Policy Q-Planning

· 50 min

Jakson kuvaus

Imitation learning, especially with interventions, has driven so much recent robotics progress. However, improving a policy via targeted interventions until it reaches a useful and deployable success rate is a time and labor intensive process. Instead, wouldn’t it be great if policies could improve on their own?
That’s what Varun Giridhar and Animesh Garg join us to talk about. In Q-Planning, they start with a large policy like pi-0.5, and add a Q-function estimator to predict value instead of just actions, then use both successful and failed rollouts to update this Q-function online, then…

Sivusto: robopapers.substack.com

Livenä, viimeisin tunti: 10 907 kappaletta 4 929 asemalla 138 maassa

Löydä maailma radion kautta

Lisää tapoja löytää (7)

Kuulitko jotain hyvää? Tallenna se tai avaa se suoraan Spotifyssa.

★ Artist of the week The fastest riser on the radio this week
Your song here? From € 9 per week →
Radio WNCS 104.7 FMJatka kuuntelua