benchmarking novelty: models that 'feel different.'
deriving patterns from the results
finishing the first medium-sized run on multiple llm models
initializing the early idea of making it work
start searching
enter keywords to search articles.