Academic & PhD Research Intern

Role overview Conduct applied research across AI and systems, working on topics such as foundation models, post-training, autonomous reasoning, inference compilers, cryptography, and agent infrastructure. The work combines research experiments with implementation and may lead to open-source releases, reproducible benchmarks, or academic publications.

Responsibilities - Investigate research questions in model architecture, RLHF/RLAIF post-training, agentic reasoning, or inference compilation. - Run large-scale experiments using provided GPU compute, API credits, and research workspaces. - Develop and share open-source model weights, benchmarks, and research results; pursue peer-reviewed publication where appropriate. - Work with researchers and engineers to integrate promising work into production open-source codebases.

Requirements - Currently pursuing a PhD, master’s, or exceptional undergraduate degree in computer science, machine learning, physics, mathematics, or a related field. - Strong implementation ability in one or more of Python, PyTorch, Rust, C++, or Go. - Evidence of research or systems work, such as open-source contributions, preprints, or substantial projects.

Nice to have - First-author papers or preprints in machine learning, compilers, cryptography, or distributed systems. - Research experience aligned with foundation models, inference systems, or related areas.

Benefits and work setup The internship or fellowship is offered remotely or in San Francisco. Researchers receive compute and API credits and access to collaborative workspaces for their lab or research group. The role supports publishing research under the researcher’s own name and offers direct collaboration with research and systems teams.