Postdoc: Mechanistic AI Interpretability & Learning Dynamics

Artois University; France; Vor Ort; EUR 42.000 - 56.000; Vollzeit

Zusammenfassung

Artois University seeks a First Stage Researcher (R1) for a temporary postdoc in the Computer Science field. The project focuses on interpreting internal mechanisms of AI models, linking internal weights and activations to human-interpretable concepts to enhance transparency and reliability.

The role requires a PhD in Computer Science, AI, or ML, with strong deep learning and explainable AI skills, and proficiency in Python and PyTorch.

Qualifikationen

• PhD in Computer Science, AI, or ML is required.
• Strong skills in deep learning and explainable AI.
• Proficiency in Python and PyTorch or equivalent frameworks.

Aufgaben

• Contribute to the task entitled “Interpreting the Internal Mechanisms of Models” by analysing AI models and their internal representations.
• Map and understand internal representations across modalities, disentangling latent concepts.
• Analyse learning dynamics and design architectures that promote performance and interpretability.
• Apply mechanistic understanding to reduce hallucinations in AI systems.

Ausbildung

PhD in Computer Science, AI, or Machine Learning

Tools

PyTorch

Jobbeschreibung

Artois University seeks a First Stage Researcher (R1) for a temporary postdoc in the Computer Science field. The project focuses on interpreting internal mechanisms of AI models, linking internal weights and activations to human-interpretable concepts to enhance transparency and reliability.

The role requires a PhD in Computer Science, AI, or ML, with strong deep learning and explainable AI skills, and proficiency in Python and PyTorch.