Dorsa Sadigh

About

Dorsa Sadigh is a leading AI and robotics researcher whose work spans human-robot interaction, autonomous systems, and large-scale machine learning. Based at Stanford University, she has made foundational contributions to how autonomous agents—particularly self-driving cars and robotic manipulators—reason about and adapt to human behavior. Her early work challenged the conventional assumption that autonomous vehicles should merely predict and avoid other drivers, instead demonstrating that robots can strategically influence human actions to achieve safer, more efficient outcomes (425 citations). This insight extended into active preference learning, where she developed methods for robots to efficiently elicit and internalize human preferences through intelligently chosen queries (259 citations). Sadigh also contributed to the landmark "Foundation Models" report (2,177 citations), one of the most influential AI documents of recent years, helping shape how the research community thinks about large-scale pretrained models. More recently, she has pushed the boundaries of embodied AI through major collaborative efforts including Open X-Embodiment and DROID, building diverse robotic datasets to train generalizable manipulation policies. Her work on SpatialVLM (163 citations) further advances vision-language models with three-dimensional spatial reasoning. Across these threads, Sadigh's research consistently bridges theoretical rigor with real-world applicability, making her a defining voice in modern interactive and embodied AI.

Research Focus

Key Achievements

Top Papers

1. 1
   On the Opportunities and Risks of Foundation Models

   2,177 citations · 2021
2. 2
   Planning for Autonomous Cars that Leverage Effects on Human Actions

   425 citations · 2016
3. 3
   Active Preference-Based Learning of Reward Functions

   259 citations · 2017
4. 4
   Planning for cars that coordinate with people: leveraging effects on human actions for planning and active information gathering over human internal state

   180 citations · 2018
5. 5
   Information gathering actions over human internal state

   165 citations · 2016
6. 6
   SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilities

   163 citations · 2024
7. 7
   Safe Control under Uncertainty with Probabilistic Signal Temporal Logic

   140 citations · 2016
8. 8
   Open X-Embodiment: Robotic Learning Datasets and RT-X Models : Open X-Embodiment Collaboration⁰

   119 citations · 2024
9. 9
   DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset

   108 citations · 2024
10. 10
    Open X-Embodiment: Robotic Learning Datasets and RT-X Models

    101 citations · 2023

Key Collaborators

Contact & Links