Leads the team this role joins
Runs the team day to day
Decides on early hires
From your own Gmail and LinkedIn.
About the role
This is a full-time remote research role at Base Labs, the research group of Baseten, focused on reinforcement learning environments for language models. It suits researchers and engineers who can design tasks and rewards, run RL experiments and write the code to build and inspect environments, and it is open at several levels. The job is based in San Francisco and New York, and is listed as fully remote.
What you’ll do
- Identify capabilities and failure modes worth studying, such as software development, tool use, scientific reasoning and complex instruction following.
- Design task distributions, interactions, rewards and verifiers together, and confirm that tasks are solvable.
- Investigate shortcuts, ambiguous instructions and weak grading.
- Build methods that create environments at scale while keeping quality high.
- Run RL experiments, compare curricula, task mixtures and reward designs, and test transfer to unseen tasks.
- Develop evaluation methods that separate real capability differences from noise and grader errors.
- Work with safety researchers on environments that reveal reward hacking and misalignment.
- Publish environments, datasets, training recipes and findings that other teams can reuse.
What they’re looking for
- Evidence of choosing a useful research question, designing an informative experiment and reasoning carefully about results.
- Depth in an area such as reinforcement learning, model evaluation, agent systems, synthetic data or experimental methodology.
- Ability to write the code needed to build environments, run experiments and inspect failures.
Nice to have
- Experience with LLM training.
- Experience building something other researchers use.
- A paper, an open-source project or a strong experimental investigation as evidence of work.
What’s on offer
- Competitive compensation with meaningful equity for U.S. roles.
- 100% coverage of medical, dental and vision insurance for employees and dependents.
- Flexible PTO, paid parental leave and a company-facilitated 401(k).
- A fertility and family-building stipend for U.S. roles.
- Exposure to ML startups and networking opportunities.
Skills
Questions about this role
Is this role remote?
Yes, it is listed as fully remote, with locations in San Francisco and New York.
Do I need a PhD?
No, a PhD or a previous Research Scientist title is not required.
Is equity offered?
Yes, meaningful equity is offered for U.S. roles.
What kind of work is this?
The role centres on designing RL environments, running experiments and publishing research and tools.