Student Researcher Seed LLM - Scientific Coding Agent; LLMs × Physics​/Chemistry; RL

Student Researcher [Seed LLM - Scientific Coding Agent (LLMs × Physics/Chemistry; RL preferred)] – 2026 Start (PhD)

Location:

San Jose

Team:
Technology

Employment Type:

Intern

Job Code: A87738

• Benchmarking for scientific coding (core):
  Design a principled benchmark suite spanning tasks with literature-grounded coding and issues solving for existing codebase.
• Curate tasks/datasets and write robust unit tests with compliance of scientific knowledge.
• Implement evaluation metrics and build a clean, extensible evaluation pipeline with baselines.
• Language evolution for new scientific discovery (core):
  Prototype loops where agents propose/refine scientific language and tie these to executable code & simulations.
• Build workflows (multi-agents, planning, tool-use, self‑critique), integrate code execution sandboxes, retrieval, and experiment runners.
• Compare LLMs and inference strategies; run ablations and produce clean research artifacts (plots, tables, write-ups).

• Minimum Qualifications: Currently pursuing a PhD in Computer Science, Machine Learning, Programming Systems, Physics, Chemistry, Biology, Applied Math, or related field.
• Strong Python and software engineering:
  Git, testing (pytest), packaging, comfort with Linux/containers.
• Working knowledge of LLMs (prompting, fine‑tuning or adapters, evaluation) and at least one ML framework (eg. PyTorch).
• Solid foundations in physics and/or chemistry (e.g., classical/quantum/thermo/stat mech; physical chemistry, molecular modeling) and numerical methods (ODE/PDE, optimization, linear algebra).
• Preferred Qualifications: Reinforcement Learning (PPO/DPO, reward modeling, curriculum learning).