Generalist - GEN-1.5: Embodied Foundation Models are One-Shot Learners

Research

August 19, 2026

Generalist Team

17 min read

Listen

Table of Contents

Introduction

Introducing GEN-1.5

Scaling Pretraining for Robotics

One-Shot Learning In-Context

Compositional Generalization

Zero-Shot Sim-to-Real Transfer

Human-to-Robot In-Context Learning

Few Gradient Step Adaptation

Physical Generalization

Looking Ahead

Citation

Back to Blog

GEN-1.5<br>Embodied Foundation Models are One-Shot Learners

Humans have a remarkable ability to perform new physical skills from only one or a few examples. Our latest robot foundation model, GEN-1.5, exhibits the beginnings of that same ability: it can learn a new task in seconds, from a single example, without gradient updates or fine-tuning. It displays broad capabilities across one-shot and few-shot learning from demonstration, as well as zero-shot physical generalization. Although the tasks are simple and short-horizon, this is the first model we know for which one-shot and few-shot learning of physical skills have emerged at scale. We view these results as a significant step towards our mission of building general intelligence for the physical world.

The promise of a robot foundation model is simple to state: walk up to a robot, and get it to do any task almost immediately. Whether zero-shot, one-shot, or few-shot, what matters from a capability standpoint is immediacy and generality: can the model learn a new task quickly and generalize to new situations? For physical tasks, this level of intelligence demands both broad abilities in comprehending task intent, as well as adapting in real time, closed-loop, to the unexpected variation of the real world.

For language models, the ability to quickly learn new tasks from just one or a few examples arrived as a hallmark capability of GPT-3.1 Across a broad suite of language tasks, it achieved roughly 45% average accuracy with one-shot in-context prompting without training, and up to ~65% with few-shot (~100 examples).1 Models prior to GPT-3 had shown flashes of zero-shot ability and even initial few-shot results2, but GPT-3 achieved significantly broader few-shot performance, paired with what at the time was an immense step in generalization capabilities.

In robotics, the analogous pursuit of systems that could generalize a task from one or a few demonstrations, has persisted for decades — tracing back at least to the teach-by-guiding of the 1954 Unimate patent3 and MIT&rsquo;s 1970 Copy Demo.4 A large number of prior works, including our own,5 have shown various forms of in-context learning but over a limited set of task variations, or under restrictions to particular objects, task types, or sensing modalities.6,7,8,9,10 The ability to learn closed-loop physical skills from just one or a few demonstrations, and to do so across a broad range of tasks without such restrictions, has predominantly been considered out of reach. Such an ability may also likely be underpinned by a foundation that enables other broad generalization capabilities.

Language model one-shot example

Q: Who wrote Romeo and Juliet?<br>A: William Shakespeare

Q: Who wrote War and Peace?<br>A: Leo Tolstoy

Embodied model one-shot example

Marker Into Cup<br>Pour Bolts<br>Zipper

Marker Into Cup

Pour Bolts

Zipper

Figure 1. One-shot learning in-context with language models and embodied models. In the language model example, the output of the model is highlighted in green. In our example, the prompt is a sensorimotor sequence from the human demonstration data, and GEN-1.5 controls the robot to accomplish the inferred task.

Introducing GEN-1.5

We&rsquo;ve created GEN-1.5, our latest robot foundation model that exhibits broad one-shot and few-shot learning from demonstration capabilities, as well as zero-shot generalization, e.g. improvisation and novel tool use (e.g. brush, dustpan, etc.). GEN-1.5 is a large multimodal model that processes video input (30 seconds of memory, alongside other sensor, language, and proprioceptive inputs) and produces 100 Hz action trajectories. Its capabilities include:

One-shot learning via in-context prompting. The model learns new tasks in seconds when prompted with 3 to 12 seconds of a single demonstration, no training required. We refer to the use of sensorimotor examples in the context window as &ldquo;physical prompting.&rdquo;

Compositional generalization. Given two different physical prompts in context, the model chains them into a single longer-horizon behavior.

Zero-shot sim-to-real transfer. A demonstration recorded in simulation works as a physical prompt for a real-world task, even though pretraining contains no simulation data.

Human-to-robot imitation. In some cases a person can demonstrate a task with their own hands, in view of the robot&rsquo;s cameras, and the model reproduces it with the robot&rsquo;s hands.

Few-shot adaptation via gradient descent. The model can be fine-tuned to a new task in 1–10...