Abstract: The progression from data-driven methodologies to the establishment of chemical symbols represents a fundamental process in the advancement of chemical education. The integration of logical reasoning with model construction serves as a pivotal approach for comprehending abstract chemical symbols, constituting a cornerstone of chemistry bridging education. This study employs the Transformer model to simulate and compute Avogadro’s constant, while utilizing Scikit-learn’s integrated models to deduce the most probable distribution. Through the implementation of the triple representation teaching strategy, the research investigates the application of AI models in chemistry bridging education, with particular emphasis on elucidating the conceptual underpinnings of Avogadro’s constant as a proportionality factor. This exploration extends to Boltzmann’s constant, the Boltzmann formula (interpretation of entropy), and the definition and derivation of temperature formulas. This pedagogical approach facilitates students’ mastery and comprehension of chemical symbols, enabling them to develop a profound understanding of chemistry as a discipline fundamentally concerned with the study of aggregates.

Key words: Artificial intelligence model, Chemistry bridging education, Avogadro’s constant, Boltzmann formula, Most probable distribution