Generalist releases GEN-1.5 with one-shot physical-task learning
Generalist released GEN-1.5, a robot foundation model that can infer short-horizon manipulation tasks from seconds of demonstration data and adapt with one to ten gradient updates.
Evidence notes
- Across 10 tasks, Generalist reports 59% average success from one 3–12 second in-context demonstration and 83% after 10 gradient steps on five minutes of task data.
- The multimodal model maintains 30 seconds of video context alongside sensor, language and proprioceptive inputs, and produces action trajectories at 100 Hz.
- The release also shows prompts transferring from simulation to a physical robot and, in some cases, from human-hand demonstrations to immediate robot execution.
Company context
Generalist builds embodied foundation models for robots, initially focused on dexterous manipulation. Its work combines real-world robot interaction data, in-house data-collection hardware and models intended to generalize across tasks.