Master Thesis in Visual Representation of Technical Diagrams for LLMs - remote possible

Master studies in Computer Science, Data Science, or comparable fields. Proficiency in Python. Prior experience in building RAG systems or chatbots is ideal. A self-driven, proactive, and solution-oriented individual with a strong independent work ethic. Strong passion for developing the best possible solutions. Fluent in English.

Research optimal methods for processing and representing technical diagrams for integration with LLMs. Explore challenges and limitations of current multi-modal LLMs with visual data. Design a pipeline for creating a synthetic dataset linking diagrams to textual descriptions. Investigate how dataset quality and structure influence model training and performance. Conduct testing on real-world technical diagrams from various domains. Develop metrics to assess model interpretation and generation based on diagram inputs. Perform comparative analysis of multi-modal models on handling technical diagrams. Investigate potential of abstract markup syntax for universal representation of technical diagrams.