# Severity-Aware Curriculum Learning with Multi-Model Response Selection for Medical Text Generation

arXiv: https://arxiv.org/abs/2606.05510 · HTML: https://mlsanity.com/paper/2606.05510/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 28 references that are in the map.
- First posted on arXiv 3 Jun 2026.

## Builds on (most cited first, 21 of 21)

- [Large language models in healthcare and medical domain: A review](https://mlsanity.com/paper/2401.06775/) (2023)
- [Competence-based Multimodal Curriculum Learning for Medical Report Generation](https://mlsanity.com/paper/2206.14579/) (2022)
- [CDL: Curriculum Dual Learning for Emotion-Controllable Response Generation](https://mlsanity.com/paper/2005.00329/) (2020)
- [Fine-Tuning by Curriculum Learning for Non-Autoregressive Neural Machine Translation](https://mlsanity.com/paper/1911.08717/) (2019)
- [KG-Rank: Enhancing Large Language Models for Medical QA with Knowledge Graphs and Ranking Techniques](https://mlsanity.com/paper/2403.05881/) (2024)
- [Does the Order of Training Samples Matter? Improving Neural Data-to-Text Generation with Curriculum Learning](https://mlsanity.com/paper/2102.03554/) (2021)
- [When Can We Trust LLMs in Mental Health? Large-Scale Benchmarks for Reliable LLM Evaluation](https://mlsanity.com/paper/2510.19032/) (2025)
- [DocLens: Multi-aspect Fine-grained Evaluation for Medical Text Generation](https://mlsanity.com/paper/2311.09581/) (2023)
- [Natural Language Processing for Smart Healthcare](https://mlsanity.com/paper/2110.15803/) (2021)
- [A Multi-Layered Large Language Model Framework for Disease Prediction](https://mlsanity.com/paper/2502.00063/) (2025)
- [In-sample Curriculum Learning by Sequence Completion for Natural Language Generation](https://mlsanity.com/paper/2211.11297/) (2022)
- [Evaluation of LLMs in Medical Text Summarization: The Role of Vocabulary Adaptation in High OOV Settings](https://mlsanity.com/paper/2505.21242/) (2025)
- [An Ensemble Classification Approach in A Multi-Layered Large Language Model Framework for Disease Prediction](https://mlsanity.com/paper/2509.02446/) (2025)
- [Few-Shot Optimized Framework for Hallucination Detection in Resource-Limited NLP Systems](https://mlsanity.com/paper/2501.16616/) (2025)
- [Arabic Large Language Models for Medical Text Generation](https://mlsanity.com/paper/2509.10095/) (2025)
- [Efficient Medical Question Answering with Knowledge-Augmented Question Generation](https://mlsanity.com/paper/2405.14654/) (2024)
- [TACL: Threshold-Adaptive Curriculum Learning Strategy for Enhancing Medical Text Understanding](https://mlsanity.com/paper/2510.15269/) (2025)
- [Severity-Aware Weighted Loss for Arabic Medical Text Generation](https://mlsanity.com/paper/2604.06346/) (2026)
- [A Severity-Based Curriculum Learning Strategy for Arabic Medical Text Generation](https://mlsanity.com/paper/2604.06365/) (2026)
- [Towards Conditioning Clinical Text Generation for User Control](https://mlsanity.com/paper/2502.17571/) (2025)
- [Scaling Arabic Medical Chatbots Using Synthetic Data: Enhancing Generative AI with Synthetic Patient Records](https://mlsanity.com/paper/2509.10108/) (2025)

## Abstract (arXiv)

Telehealth systems have become increasingly important for delivering accessible and timely medical information. Existing large language models often struggle to provide consistent and contextually appropriate medical responses across varying levels of case severity. This limitation highlights the need for models that can effectively adapt to the progressive complexity in medical queries. To address this challenge, we introduce a severity-aware multi-model framework that integrates curriculum training strategy with relevance-based response selection. The proposed framework employs a three-stage curriculum learning strategy, where each model is trained sequentially on mild, moderate, and critical cases to progressively acquire domain knowledge. The approach uses five large language models, each trained independently under the same curriculum. During inference, all models generate candidate responses, and the response with highest BERTScore is selected as the final output. The framework is trained and evaluated on the MAQA dataset, which provides annotated medical question-answer pairs. Experimental results evaluated using BERTScore demonstrate that the proposed method achieves superior performance compared to both baseline and fine-tuned models, attaining 86.71% in the baseline setting and 90.30% after fine-tuning. These results highlight the effectiveness of combining curriculum learning with multi-model response selection in improving response quality and relevance in medical text generation.