Member of Technical Staff, Reinforcement Learning

Baseline Robotics; Zürich; Vor Ort; CHF 120.000 - 145.000; Vollzeit

Benefits dieser Stelle

Stock options

Zusammenfassung

Baseline Robotics in Zürich, Switzerland, seeks a researcher to advance reinforcement-learning systems on real hardware. You will run post-training loops, design reward models, and implement on-robot safety constraints to prevent hardware damage.

The role emphasizes translating simulation insights to real outcomes, aligning policy improvements with measured success, and collaborating across ML and robotics teams. On-site, full-time with stock options.

Qualifikationen

• Experience training reinforcement-learning systems in robotics, simulation, games, or agents.
• Knowledge of policy optimisation and offline or off-policy reinforcement learning.
• Publications in top machine learning or robotics venues, or equivalent experience.
• Fluency in Python and PyTorch, and the ability to debug a learning system end-to-end.
• Experience with JAX or Rust.
• Experience with GPU-accelerated simulation tools (Isaac Lab, Newton, mjlab, Genesis).

Aufgaben

• Run post-training and reinforcement-learning loops for on-robot policy improvement.
• Design reward models from vision, language, and operator feedback.
• Implement on-robot safety constraints to halt or recover a trial before hardware damage.
• Post-train a reference policy and account for the measured change in success rate.
• Reconcile imagined or simulated rollouts with hardware outcomes when transfer is poor.

Kenntnisse

Reinforcement learning

Python

PyTorch

Publications in ML/robotics

JAX

Rust

GPU simulation

Tools

Isaac Lab

Genesis

mjlab

Newton

Jobbeschreibung

About Baseline

Open source revolutionised how we write software, but its impact on robotics stays limited due to expensive hardware accessible only to well-funded labs. We run that hardware as shared infrastructure for the broader community, billed only for active compute. For a given task, policies are evaluated under similar conditions, giving researchers common ground for fair comparisons.

In this role you will

• Run post-training and reinforcement-learning loops for on-robot policy improvement.
• Design reward models from vision, language, and operator feedback.
• Implement on-robot safety constraints to halt or recover a trial before hardware damage.
• Post-train a reference policy and account for the measured change in success rate.
• Reconcile imagined or simulated rollouts with hardware outcomes when transfer is poor.

What we hope you'll bring

• Experience training reinforcement-learning systems in robotics, simulation, games, or agents.
• Knowledge of policy optimisation and offline or off-policy reinforcement learning.
• Publications in top machine learning or robotics venues, or equivalent experience.
• Fluency in Python and PyTorch, and the ability to debug a learning system end-to-end.
• Experience with JAX or Rust.
• Experience with GPU-accelerated simulation tools (Isaac Lab, Newton, mjlab, Genesis).

Details

• Location: Zurich, Switzerland, on-site 5 days a week.
• Annual salary: CHF 120,000 - 145,000.
• Every full-time hire receives stock options as part of the stated salary.

Deadline to apply: None. Applications will be reviewed on a rolling basis.