Is it possible to extract the internal physics knowledge of a large language model as semantic network?ΒΆ

The answer of this question depends on the answer of the following questions:

The following approaches can be used for extracting the physics knowledge from a LLM:

  • Building a subtopic network