[ICLR 2025] Agent S: an open agentic framework that uses computers like a human
Awesome In-Context RL: A curated list of In-Context Reinforcement Learning - - —
🧠🔗 Graph-Based Programmable Neuro-Symbolic LM Framework - a production-first LM framework built with decade old Deep Learning best practices
Official Implementation for "In-Context Reinforcement Learning for Variable Action Spaces"
#数据仓库#XLand-100B: A Large-Scale Multi-Task Dataset for In-Context Reinforcement Learning - - —
Vintix: Action Model via In-Context Reinforcement Learning - - —
Official Implementation for "In-Context Reinforcement Learning from Noise Distillation"
Xenoverse is a collection of randomized RL, Language, and general-purpose simulation environments, designed for training General-Purpose Learning Agents (GLAs).
Next-gen Foundation Model for Embodied AI
Toy meta-RL environments for testing algorithms implementations
Repository Switched To Xenoverse (https://github.com/FutureAGI/Xenoverse)