This dataset augments BoltzmannEntropy/QuantumLLMInstruct with a new field subject_knowledge,
automatically generated from each example’s main_domain, sub_domain, and problem using GPT-5.
train.json (rows: 5150): Preserves original fields and adds:
subject_knowledge (string): 3–6 concise lines summarizing definitions, governing equations,
assumptions/scales, and a typical solution strategy relevant to the… See the full description on the dataset page:
https://huggingface.co/datasets/kalkiai3000/quantum-llm-instruct-subject-knowledge.
