Technical Staff Member, Research

This is a fully remote position, open to applicants in Canada.

📋 Description

• Conduct research, design, and evaluate innovative model architectures specific to research that incorporate academic literature, natural language processing (NLP), symbolic reasoning, and various methods to facilitate the scientific process.

• Create and implement custom tokenizers for LaTeX symbols and physical units to be recognized as tokens.

• Investigate alternatives to transformers through comprehensive research and offer actionable recommendations for model advancement.

• Design reinforcement-learning loops that empower models to conduct independent and internal thought experiments.

• Collaborate with our Data Scientists & Engineers to design and automate data ingestion pipelines that compile scientific literature, metadata, experimental data, equations, and additional data sources in a reliable and scalable way.

• Create custom benchmarks to evaluate the models’ comprehension of physical concepts, mathematical reasoning capabilities, and their ability to reduce hallucinations, enhancing scientific reliability.

• Enhance and publish datasets and baselines after ensuring internal tests are stable.

• Oversee and monitor model training jobs while guiding the technical team through setup, tracking progress, and managing costs within the established budget.

• Formulate methods to conduct “practice runs” in a sandbox environment to enhance the model’s ability to independently explore ideas while documenting results for subsequent review.

• Create a framework for assessing the models’ learning through visual and statistical tools to identify patterns and blind spots.

• Implement guardrails and tests that signal poor-quality model outputs.

• Maintain internal tools to monitor known issues, documenting failures, clear solutions, and enhancements for future development integration.

• Collaborate with the engineering team to ensure product viability and robust architecture.

• Effectively communicate technical trade-offs to non-technical stakeholders in understandable terms.

• Provide clear updates on findings to the technical team to keep the broader team informed about progress against research milestones.

⛳️ Requirements

• Educational Background: PhD in physics, computer science, data science, information systems, or a related discipline.

• Experience: Demonstrated success in conducting extensive research on scientific AI models, symbolic models, machine learning, or deep learning for scientific discovery.

• Technical Skills: Knowledge of state-of-the-art models, best practices in model development processes, comprehensive understanding of AI/ML concepts, and data infrastructure.

• Collaboration & Communication:

• Ability to work closely with engineers and other technical team members.

• Excellent written and verbal communication skills.

• Comfortable functioning within a startup-style, cross-functional, remote team.

• Bonus Skills:

• Experience with or a strong interest in physics and/or fundamental science topics.

• Experience in conducting research on AI models in an early-stage or mission-driven environment.

🏝️ Benefits

• Join us at FirstPrinciples and be part of a transformative journey where science propels progress and unlocks humanity's potential.