Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study explores the impact of using bounding boxes to crop images of handwritten student responses on the performance of Small Language Models (SLMs) in grading tasks. By evaluating models with parameter sizes from 4B to 72B, the authors find that cropping significantly enhances grading accuracy and reduces computational costs when processing complex visual data. The findings suggest that bounding boxes are essential for optimizing SLMs in educational assessments, particularly in scenarios involving large-scale image inputs.
Cropping images with bounding boxes can drastically boost Small Language Models' grading accuracy while slashing computational costs.
The deployment of Small Language Models (SLMs) in educational settings offers significant advantages in terms of privacy, cost, and scalability. However, SLMs often struggle with complex vision-based tasks, such as grading handwritten student exams, due to the high computational cost of processing large images and the visual distractions present on a full page. In this paper, we investigate whether cropping student responses using bounding boxes can improve the accuracy and computational efficiency of SLMs on a short-answer grading task. Using a dataset of scanned handwritten responses from the 2025 Australian Physics Olympiad, we evaluate the performance of several models ranging from 4B to 72B parameters under varying conditions of Chain of Thought (CoT) prompting and image cropping. Our results demonstrate that using bounding boxes significantly improves grading accuracy and reduces computational cost (FLOPs) across models. We conclude that bounding boxes are a crucial pre-processing step for deploying SLMs in large-scale, vision-based educational assessments.