"Reflexion: Language Agents with Verbal Reinforcement Learning" Noah Shinn | AI4Code Meetup

The London AI for Source Code group is an active research hub with researchers and academics interested in all applications of AI on source code. Once a month, esteemed speakers from around the world are invited to share their latest research work in our online 40-minute sessions. Get invited to our next meetup: Subscribe to our invitation list: https://ai4code-london.github.io/contact/ In this meetup, Noah Shinn (researcher @ at Northeastern University) joined us to discuss his recent work on a novel verbal Reinforcement Learning framework, created to reinforce language agents not by updating weights, but instead through linguistic feedback (https://arxiv.org/abs/2303.11366).