Papers
10
Total Citations
255
H-Index
6
About
Aniruddha Kembhavi is a leading researcher at the intersection of computer vision, natural language processing, and embodied AI, where he drives the development of generalist agents that can perceive, reason, and act in the physical world. As a key figure behind the Allen Institute for AI (AI2), his most impactful contribution is **Unified-IO 2**, the first autoregressive multimodal model capable of understanding and generating text, images, audio, and action—a foundational step toward truly unified AI systems. His work on **Open X-Embodiment** (119 citations) has been instrumental in consolidating robotic learning across diverse datasets, enabling large-scale, general-purpose robot policies. Kembhavi also created **AllenAct**, a widely adopted framework for embodied AI research (44 citations), and **ManipulaTHOR**, which advanced visual object manipulation in simulation. More recently, his **Phone2Proc** method bridges the sim-to-real gap by using a 10-minute phone scan to generate realistic training environments, while **Promptable Behaviors** personalizes robot policies to human preferences. With over 250 total citations and a consistent focus on scaling multimodal, interactive agents, Kembhavi’s work is shaping the future of robots that can navigate, manipulate, and collaborate in our chaotic, real-world environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3AllenAct: A Framework for Embodied AI Research44 citations · 2020
- 4Phone2Proc: Bringing Robust Robots into Our Chaotic World9 citations · 2023
- 5ManipulaTHOR: A Framework for Visual Object Manipulation7 citations · 2021
- 6
- 7
- 8Harmonic Mobile Manipulation3 citations · 2024
- 9Seeing the Unseen: Visual Common Sense for Semantic Placement3 citations · 2024
- 10