In recent years, large language models (LLMs) have gained significant traction across various domains, including education. This paper explores the application of LLMs in grading programming assignments. By leveraging data collected from existing programming assignments and their corresponding grades, we aim to develop a robust LLM-based grading system. We also incorporate augmented data representing various grading scenarios to enhance the model’s performance and ensure comprehensive coverage across all grading levels. Our approach involves training the LLM on this combined dataset to enable accurate and consistent evaluation of programming assignments. The proposed model, BeGrading, aims to reduce the grading burden on educators and provide timely and objective feedback to students. Compared to the Codestral model, our proposed model demonstrates an absolute difference rate of 19%, equivalent to $$\pm 0.95$$ out of 5. This is acceptable for using a small, fine-tuned model with optimized data. Additionally, the Codestral model compared to the dataset optimized score shows a difference of 15% equivalent to a margin of $$\pm 0.75$$ out of 5. Preliminary results demonstrate the potential of LLMs to perform grading tasks with a high degree of reliability, opening avenues for further research and practical applications in automated education systems.
The growing demand for radiological services, amplified by a shortage of qualified radiologists, has resulted in significant challenges in managing the increasing workload while ensuring the accuracy and timeliness of radiological reports. To address these issues, recent advancements in artificial intelligence (AI), particularly in transformer models, vision-language models (VLMs), and Large Language Models (LLMs), have emerged as promising solutions for radiology report generation (RRG). These systems aim to make diagnosis faster, reduce the workload for radiologists by handling routine tasks, and help generate high-quality, consistent reports that support better clinical decision-making. This comprehensive study covers RRG developments from 2021 to 2025, focusing on emerging transformer-based and VLMs, highlighting the key methods, architectures, and techniques employed. We examine the datasets currently available for RRG applications and the evaluation metrics commonly used to assess model performance. In addition, the study analyzes the performance of the leading models in the field, identifying the top performers and offering insights into their strengths and limitations. Finally, this study proposes new directions for future research, emphasizing potential improvements to existing systems and exploring new avenues for advancing the capabilities of AI in radiology report generation.
The rapid advancement of Large Language Models (LLMs) has led to substantial investment in enhancing their capabilities and expanding their feature sets. Despite these developments, a critical gap remains between model sophistication and their dependable deployment in real-world applications. A key concern is the inconsistency of LLM-generated outputs in production environments, which hinders scalability and reliability. In response to these challenges, we propose a novel framework that integrates custom-defined, rule-based logic to constrain and guide LLM behavior effectively. This framework enforces deterministic response boundaries while considering the model’s reasoning capabilities. Furthermore, we introduce a quantitative performance scoring mechanism that achieves an 85.5% improvement in response consistency, facilitating more predictable and accountable model outputs. The proposed system is industry-agnostic and can be generalized to any domain with a well-defined validation schema. This work contributes to the growing research on aligning LLMs with structured, operational constraints to ensure safe, robust, and scalable deployment.
6 October 2025 in Computers in Biology and Medicine