PhD Internship: LLM Interpretability

Hey there We are SPRYFOX - a young startup building AI solutions that show the value of analytics to customers across industries. As an AI boutique, we develop custom AI solutions that drive real business impact. We're looking for a PhD Intern to join our team. With your 3+ years of ML/AI research experience, you'll work on a novel LLM-based project that translates research insights into real-world systems. You genuinely enjoy the challenge of making complex technical work meaningful for business stakeholders. Our consulting setup means variety - different industries, different problems, always something new. What You'll Do Conduct research in cutting-edge fields like mechanistic interpretability Work with methods including causal mediation analysis and model editing Research and implement novel methods Help design end-to-end LLM pipelines that are interpretable and efficient Collaborate in a team and help establish technical best practices What We're Looking For Technical Experience Currently enrolled in, or recently completed, an NLP-related PhD program Strong Python skills, ML frameworks (scikit-learn, Huggingface, PyTorch) Solid LLM experience: fine-tuning, prompt design, evaluation Able to work full-time during the period of the internship Initial experience in mechanistic interpretability and familiarity with related frameworks (TransformerLens, NNSight) is a plus. Collaboration & Communication Strong scientific communication skills (writing, reading, presenting) Strong command of English for presentations and technical documentation Job Type : Full-time