AIITDec 7, 2025

FlatFormer: A Flat Transformer Knowledge Tracing Model Based on Cognitive Bias Injection

arXiv:2512.06629v11 citationsh-index: 1
Originality Incremental advance
AI Analysis

This addresses the computational inefficiency of deep models for real-time deployment in educational technology, representing a strong incremental improvement.

The paper tackled the performance-complexity trade-off in Knowledge Tracing models by proposing FlatFormer, a flat Transformer with cognitive bias injection, which achieved an 8.3% higher AUC on the EdNet dataset while using less than 15% of parameters and being three times faster than hierarchical baselines.

Knowledge Tracing (KT) models face a critical ``Performance-Complexity Trap'': capturing complex cognitive dynamics like learning sessions and memory decay typically requires deep hierarchical architectures, which incur prohibitive computational costs for real-time deployment. To resolve this, we propose FlatFormer, a streamlined architecture based on the novel design paradigm of ``Information Injection over Structural Stacking.'' Unlike parameter-heavy hierarchical models, FlatFormer leverages a standard flat Transformer augmented with two lightweight injection mechanisms: (i) a hybrid input encoding strategy combining learnable session identifiers with fixed sinusoidal step embeddings; and (ii) a pre-computed power-law bias integrated directly into attention logits to explicitly model the forgetting curve. Extensive experiments on four large-scale datasets (e.g., EdNet, Junyi) show that FlatFormer achieves state-of-the-art performance. For example, on the EdNet dataset, compared to the strongest hierarchical baseline (HiTSKT), its absolute AUC increased by 8.3%, while using less than 15% of parameters, and inference speed was about three times faster. These results validate that high cognitive fidelity does not necessitate architectural complexity.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes