Postdoctoral Research Assistant, Digital Humanities and Media Technology Studies

About the role

The position leads the computational side of a digital humanities project, developing models, attribution pipelines, experiments, and benchmarks for historical literary texts.

Key responsibilities

• Prepare and validate historical corpora, including OCR quality and metadata, together with the research assistant.
• Build and fine‑tune open‑weight language models on period texts using parameter‑efficient methods such as LoRA.
• Design and run systematic model tests, controlled and adversarial prompting experiments at scale with literary historians.
• Lead the technical release of the benchmark, including documentation, baselines, a public leaderboard, and integration with existing evaluation frameworks.
• Co‑author papers for NLP, machine‑learning, and humanities venues and participate in workshops and outreach activities.

Required profile

• Doctorate (by appointment) in computer science, NLP, computational linguistics, machine learning, digital humanities or a related field with a strong computational component.
• Strong interdisciplinary interest in humanities, literary theory, or philosophy of language.
• Excellent English communication skills; German is a plus.
• Residence in the Basel trinational region for the project duration.

Required skills

• Hands‑on experience with Python, PyTorch and the Hugging Face ecosystem for training and fine‑tuning transformer language models.
• Experience working on GPU clusters.
• Proficiency with containerised environments such as Docker or Apptainer and version‑control best practices.
• Familiarity with interpretability methods (e.g., feature attribution, probing, SHAP) and benchmark creation.