[BA] Local Explanations for Bigraph Programs: Aiding Model Comprehension

Ensuring that AI-generated outputs comply with policy-based constraints is a key challenge in AI. This thesis explores whether large language models can explain and help correct failures in rule-based reactive system models, such as bigraphs. By generating natural language explanations for rule application failures, the work aims to support the development of more interpretable and policy-aligned program variants.

Betreuer: Dominik Grzelak