Trusting Language Models in Education
This addresses the practical issue of unreliable AI in education, though it appears incremental as it builds on existing calibration methods with a specific feature approach.
The paper tackles the problem of language models making errors in educational question-answering by proposing to calibrate their confidence using attention-based features with XGBoost on BERT, resulting in improved probability estimates to avoid misleading students.
Language Models are being widely used in Education. Even though modern deep learning models achieve very good performance on question-answering tasks, sometimes they make errors. To avoid misleading students by showing wrong answers, it is important to calibrate the confidence - that is, the prediction probability - of these models. In our work, we propose to use an XGBoost on top of BERT to output the corrected probabilities, using features based on the attention mechanism. Our hypothesis is that the level of uncertainty contained in the flow of attention is related to the quality of the model's response itself.