Generalist's GEN-1.5 learns new robot tasks from seconds of demos, zero fine-tuning
量子位 · wechat · 2026-08-21
Generalist AI released GEN-1.5, a robot foundation model with one-shot in-context learning: show it a 3-12 second demonstration (a "physical prompt") and it performs the new task with zero gradient updates or fine-tuning, in roughly half a minute. It can also chain two separate demos into one continuous task and transfer a simulator demo straight to the real world.
Notably, the ability wasn't explicitly engineered—no special architecture, no meta-learning loop—it emerged from large-scale pretraining on continuous real-world physical data, echoing how in-context learning emerged in GPT-3. During pretraining the team saw required fine-tuning data shrink from hundreds of steps to tens, then 10, then zero.
Benchmarks: across 10 tasks, 59% average success with zero gradient updates; 83% with 5 minutes of extra data and 10 gradient steps per task. Current tasks are mostly short-horizon and stability still lags full fine-tuning, but 59% with no training is what excites researchers.
Related event: Generalist AI Unveils GEN-1.5, a One-Shot Learner for Robots(19 posts)→
More from Embodied
- EDC Wearables Stack: Ultrahuman Ring, Google Health, and IDUN EEG Earbuds — melnykowycz · 2026-08-21
- California startup LUUM built a robot that applies 25 lash extensions per eye — lukas_m_ziegler · 2026-08-21
- Building an offline talking Halloween prop: Stack and challenges — breksyt · 2026-08-21
- Prediction: Cognitive AGI, Autonomous Driving, and Robots Arriving by End of Year — davidpattersonx · 2026-08-21
- VLM Agent for Humanoids Performs Tasks and Conversation — Kangwook_Lee · 2026-08-21
- AdaPT Framework Lets Robots Mimic Federer's Playing Style — Scobleizer · 2026-08-21