IRIS
$ 0
Imagination-based agent using a transformer world model for RL; strong sample efficiency in simulated environments
Available on backorder
Brain Score
3
Specifications and details:
| Nationality | US |
|---|---|
| Website | https://iris.ai/ |
| Model type | Foundation Model |
| Manufacturer | IRIS AI |
| Release date | 2022 |
Description
IRIS is a foundation model that enables robots to plan actions by imagining outcomes before execution. Instead of relying solely on trial and error, it builds an internal model of how environments respond to different actions. Consequently, robots can evaluate strategies mentally and select more effective behaviors. Moreover, this imagination-driven approach reduces dependence on extensive real-world testing, which accelerates learning and iteration.
In addition, IRIS emphasizes efficiency and structured decision-making. Therefore, it achieves strong performance with fewer interactions compared to traditional methods. Its transformer-based design allows the model to understand sequences of actions and predict their effects over time. Because it trains effectively in simulated environments, developers can refine robot skills quickly and safely. As a result, supports faster progress toward intelligent, adaptable, and general-purpose robotic behavior.
New! 2026 Humanoid
Robot Market Report
198 pages of exclusive insight from global robotics experts — uncover funding trends, technology challenges, leading manufacturers, supply chain shifts, and surveys and forecasts on future humanoid applications.
now Google DeepMind
Contact Humanoid.guide
Website: https://iris.ai/
