Stopping Agents

Language agents for optimal stopping

Stopping Agents is an open-source framework for developing language agents that learn to optimally stop conversations. It provides a custom gym-style environment instantiated from conversation transcripts along with state-of-the-art algorithm implementations for an easy entry point to research and development.

Made with ❤ by Emaad Manzoor, Eva Ascarza, & Oded Netzer.