A structured curriculum for teaching multiplication to small language models. This dataset moves from rote memorization to logical decomposition.
Tier 1: Basic Tables (1-100) - Sequential and shuffled pairs for results up to 1000 to establish "Instant Recall" weights.
Tier 2: Sign Logic - Explicitly teaches the rules of arithmetic signs: $(+) \times (+)$, $(-) \times (+)$, and $(-) \times… See the full description on the dataset page:
https://huggingface.co/datasets/shreeman-iyer/llm_multiplication_training.
