Graph Representation Insights
Gilles discusses the expectations surrounding large language models and their ability to represent complex community structures, like that of the Karate Club graph. Kyle emphasizes the importance of recognizing whether the generated graphs maintain recognizable features rather than exact replicas. Erwan highlights the varying performance of different LLMs, noting that while some produce reliable representations, others may struggle with accuracy.In this clip
From this podcast
Data Skeptic
Auditing LLMs and Twitter
Related Questions