Applied Research - Continual Learning
Core
Build continual learning systems that convert agent trajectories and failures into model updates and persistent memory for frontier-scale AI.
Role type
Applied Research Scientist (Continual Learning & RL)
Builds
Open superintelligence stack, post-training systems, and agent workflows for autonomous research and long-horizon tasks.
Domain
Frontier AI, Reinforcement Learning, Continual Learning, Infrastructure
Deliverable
production ML models
Required skills
Continual/online learning, Reinforcement Learning (RL), Meta-learning, Memory systems, Post-training data loops, Adaptive production ML, Python, Experimental design, Model debugging
Preferred skills
Publications, Open-source contributions, Benchmarks, Shipped products
Technologies
Python, RL frameworks, Model training pipelines, Evaluation harnesses
Responsibilities
Run post-training experiments against real baselines to determine model vs. data gains; Build task-generation and training pipelines with reproducible data; Design evaluations for retention, transfer, and regression; Implement promotion gates and human oversight for changing agents.
Seniority
Senior, hands-on IC
