Research Scientist
Pluralis Research, San Francisco +1
What you will do
- Solve the open problems: Identify the questions that block Protocol Learning at scale. communication efficiency, convergence under churn and staleness, heterogeneity, robustness to malicious participants.
- Publish in Tier-1 venues: The problems are hard and largely unclaimed. Solve them, and publish.
- Get the methods into runs: Work with the engineering team so your results land in live training runs, not just papers.
What you need
- Research excellence (required): PhD in machine learning with publications in top-tier conferences (NeurIPS, ICML, ICLR).
- Distributed ML experience: Hands-on experience in large-scale distributed training and compression strategies.
- Implementation skills: Strong programming ability in PyTorch.