Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will conduct research on the thermodynamic cost of distributed computation with the principal investigator. You will develop mismatch-cost theory and approximation techniques for large state spaces, apply them to neural networks and language models, investigate computational complexity, and contribute through publications and workshop organization.
Requirements
- Ph.D. in Physics, Computer Science, or a related field by the start date
- Background in stochastic thermodynamics, statistical physics, or theoretical computer science
- Familiarity with Monte Carlo methods, uncertainty quantification, backward differential equations, or tensor network approximations
- Strong programming skills
- Legal ability to work in the United States
Responsibilities
- Collaborate with the principal investigator and team members on research, publications, and workshop organization
- Develop and apply mismatch-cost theory and large-state-space approximation techniques to distributed computational systems
- Design, train, and analyze feedforward neural networks, restricted Boltzmann machines, and lightweight LLMs
- Relate network topology to thermodynamic cost, robustness, speed, and computational power
- Investigate the relationship between computational-problem hardness and thermodynamic cost