Search papers, labs, and topics across Lattice.
This paper introduces the Personalized Thinking Model (PTM), a hierarchical representation of learner cognition built from journal entries using LLMs (Gemini 2.5 Pro), embeddings, dimensionality reduction, and clustering. The PTM aims to clone a learner's thinking model across five layers: behavioral instances, patterns, cognitive routines, metacognitive tendencies, and self-system values. Experiments with 40 participants showed that the PTM achieves acceptable fidelity, as measured by automatic information point matching (F1=75.48% after HITL), user ratings (4.3/5 after HITL), and semantic coherence across layers.
LLMs can construct interpretable, multi-layered models of individual student cognition from journal entries, opening new possibilities for personalized education.
This paper presents the Personalized Thinking Model (PTM), a hierarchical and interpretable learner representation designed for AI supported education. PTM organizes evidence from learner journals into a five-layer structure covering behavioral instances, behavioral patterns, cognitive routines, metacognitive tendencies, and self-system values. PTM is grounded in Marzano's New Taxonomy of Educational Objectives and tries to clone learner's thinking model and build cognitive twin. It was constructed using a pipeline that combines large language model inference (Gemini 2.5 Pro), sentence embeddings, dimensionality reduction, and consensus clustering. This paper evaluates PTM fidelity through three methods applied to 40 participants in a seven-week study. First, automatic evaluation using atomic information point matching yielded an overall F1 score of 74.57% before human-in-the-loop (HITL) refinement and 75.48% after refinement. Second, user evaluation using a Likert scale produced mean ratings of 4.26 and 4.30 on a five-point scale for pre and post-HITL conditions respectively. Third, semantic alignment verification showed that topic coherence increased from 0.436 at the behavioral layer to 0.626 at the core value layer, while lexical overlap with journal vocabulary decreased from 0.114 to 0.007 across those same layers. These results suggest that the PTM produces outputs with acceptable fidelity, was generally perceived by users as reflecting their thinking, and showed a pattern consistent with semantic abstraction across layers.