Reinforcement Learning from Human Feedback: Progress and Challenges

Registrations are closed

Please present your e-ticket at check-in. This will ensure you have access to the auditorium and a seat.

Reinforcement Learning from Human Feedback: Progress and Challenges

A Distinguished Lecture by John Schulman from OpenAI on the Reinforcement Learning from Human Feedback (RLHF) work powering ChatGPT.

By External Relations Group, EECS

Date and time

Wednesday, April 19, 2023 · 5 - 6pm PDT

Location

Banatao Auditorium

310 Sutardja Dai Hall UC Berkeley Berkeley, CA 94720

About this event

Speaker Biography:

John received a PhD in the EECS department at Berkeley in 2016, advised by Pieter Abbeel. He now leads a team working on ChatGPT and RL from Human Feedback at OpenAI, where he was a cofounder. His previous work includes foundational algorithms of deep RL (TRPO, PPO), generalization in RL (ProcGen), mathematical reasoning by language models (GSM8K), combining RL with retrieval (WebGPT) and studying scaling laws of RL and alignment.

In his free time, he enjoys running, jazz piano, and raising chickens.

Please note: if you would like to watch John's talk remotely, you need not register. Tickets here are for in-person attendance only, for which there is very limited seating.

You will find the talk streaming here: https://www.youtube.com/c/UCBerkeleyEECSEvents/live

Sales Ended