Stanford Intelligent and Interactive Autonomous Systems Group (ILIAD) develops algorithms for AI agents that safely and reliably interact with people. Our mission is to develop theoretical foundations for human-robot and human-AI interaction. Our group is focused on: 1) formalizing interaction and developing new learning and control algorithms for interactive systems inspired by tools and techniques from game theory, cognitive science, optimization, and representation learning, and 2) developing practical robotics algorithms that enable robots to safely and seamlessly coordinate, collaborate, compete, or influence humans.
Recent News
May 13, 2024: |
Our 8 papers got accepted to the Robotics: Science and Systems (RSS), 2024: - Efficient Data Collection for Robotic Manipulation via Compositional Generalization - Imitation Bootstrapped Reinforcement Learning - Explore until Confident: Efficient Exploration for Embodied Question Answering - RT-H: Action Hierarchies Using Language - Pushing the Limits of Cross-Embodiment Learning for Manipulation and Navigation - DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset - Octo: An Open-Source Generalist Robot Policy - Learning to Learn Faster from Human Feedback with Language Model Predictive Control |
May 1, 2024: |
Our 2 papers got accepted to the International Conference on Machine Learning (ICML), 2024: - Prismatic VLMs: Investigating the Design Space of Visually-Conditioned Language Models - Chain of Code: Reasoning with a Language Model-Augmented Code Emulator |
Feb 15, 2024: |
Our 5 papers got accepted to the International Conference on Robotics and Automation (ICRA) 2024: - Physically Grounded Vision-Language Models for Robotic Manipulation - Toward Grounded Commonsense Reasoning - Distilling and Retrieving Generalizable Knowledge for Robot Manipulation via Language Corrections - How to Prompt Your Robot: A Prompt Book for Manipulation Skills with Code as Policies - Open X-Embodiment: Robotic Learning Datasets and RT-X Models |
Jan 30, 2024: | Our paper titled "Contrastive Preference Learning: Learning from Human Feedback without RL" got accepted to the International Conference on Learning Representations (ICLR) 2024! |
Sep 21, 2023: |
Our 4 papers got accepted to the Conference on Neural Information Processing Systems (NeurIPS) 2023: - Parallel Sampling of Diffusion Models (Spotlight) - Diverse Conventions for Human-AI Collaboration - Data Quality in Imitation Learning - Inverse Preference Learning: Preference-based RL without a Reward Function |
See All |