Interpretability Researcher

About the role

TAIRC is seeking an Interpretability Researcher to advance methods that make AI decisions transparent and understandable. You will work on cutting‑edge experiments, develop open‑source tools, and contribute to the scientific community.

Key responsibilities

• Plan and run experiments to uncover how models process information and generate predictions.
• Develop and maintain open‑source interpretability libraries and visualisation tools.
• Review the latest academic literature, synthesize findings, and produce internal reports.
• Conduct mathematical research on inductive biases and underlying principles of neural networks.
• Communicate results through papers, presentations, and collaboration with volunteers or partners.

Required profile

• Doctoral degree in machine learning, applied mathematics, or a closely related field.
• Proven experience in interpretability research.
• Familiarity with interpretability tools such as Integrated Gradients, SHAP, or attention visualisations.

Required skills

• Integrated Gradients
• SHAP
• Attention visualisations

What we offer

• Opportunity to shape the future of explainable AI.
• Collaboration with a dedicated research team.
• Potential for future compensated positions pending funding.