Edison Scientific
Member of Technical Staff, AI Research Scientist
$210K–$350K
San Francisco · On-site · Staff · Full-time · Posted Jun 25, 2026
- Experience
- 5-10+ years
About the job
Develops AI models and algorithms for scientific research, including LLM reasoning and sampling methods, and takes research projects from problem scoping through deployment.
What you'll do
- Research post-training of LLMs for scientific reasoning and optimal sampling techniques.
- Translate scientific-agent challenges into statistical problems and lead research projects through deployment.
- Develop and extend the experimentation platform; collaborate with researchers, engineers, biologists, and chemists.
What they ask for
- 5-10+ years of ML research on real-world problems, including areas such as reinforcement learning, representation learning, or Bayesian methods.
- Fluency in PyTorch, Jax, or an equivalent framework.
- Experience with large-scale training and experimentation technologies, such as Ray, verl, or Megatron-LM.
- Experience with data and ML systems, production validation, and experimentation in academic or industry settings; strong programming expertise.
Nice to have
- PhD in Machine Learning, Computer Science, or another quantitative field.
- Familiarity with leveraging and managing distributed systems.
Pay
$210,000 - $350,000 USD base salary
Tools
PyTorch, Jax, Ray, verl, Megatron-LM
AI focus
Research and develop AI scientist agents, including LLM post-training for scientific reasoning and optimal sampling, to accelerate scientific research.
Model-written summary · the original posting is authoritative