- Design and test novel model architectures combining NLP, symbolic reasoning, and scientific workflows.
- Develop embedding representations for physical concepts, equations, and mathematical objects.
- Investigate and recommend alternatives to transformer-based architectures.
- Design and execute experiments to guide next-generation architecture development.
- Build reinforcement learning loops for autonomous internal thought experiments.
- Automate data ingestion pipelines for scientific literature and experimental data.
- Create benchmarks to measure physical understanding and mathematical reasoning.
- Manage model training jobs, compute usage, and sandbox environments.
- Build evaluation frameworks to identify model strengths and failure modes.
- Communicate technical trade-offs and present research updates.
Machine LearningNLP