Reason2Decide: Rationale-Driven Multi-Task Learning

TL;DR

Reason2Decide is a two-stage training framework for clinical decision support that improves prediction accuracy and explanation quality, using LLM-generated rationales and reducing reliance on human annotations.

Contribution

It introduces a novel two-stage training method with scheduled sampling to enhance rationale alignment and performance in multi-task learning for clinical NLP tasks.

Findings

Outperforms fine-tuning baselines and some zero-shot LLMs in prediction and rationale fidelity.

Achieves robustness across different rationale sources, including LLM-generated and nurse-authored.

Operates effectively with models 40x smaller than large foundation models.

Abstract

Despite the wide adoption of Large Language Models (LLM)s, clinical decision support systems face a critical challenge: achieving high predictive accuracy while generating explanations aligned with the predictions. Current approaches suffer from exposure bias leading to misaligned explanations. We propose Reason2Decide, a two-stage training framework that addresses key challenges in self-rationalization, including exposure bias and task separation. In Stage-1, our model is trained on rationale generation, while in Stage-2, we jointly train on label prediction and rationale generation, applying scheduled sampling to gradually transition from conditioning on gold labels to model predictions. We evaluate Reason2Decide on three medical datasets, including a proprietary triage dataset and public biomedical QA datasets. Across model sizes, Reason2Decide outperforms other fine-tuning baselines and some zero-shot LLMs in prediction (F1) and rationale fidelity (BERTScore, BLEU, LLM-as-a-Judge). In triage, Reason2Decide is rationale source-robust across LLM-generated, nurse-authored, and nurse-post-processed rationales. In our experiments, while using only LLM-generated rationales in Stage-1, Reason2Decide outperforms other fine-tuning variants. This indicates that LLM-generated rationales are suitable for pretraining models, reducing reliance on human annotations. Remarkably, Reason2Decide achieves these gains with models 40x smaller than contemporary foundation models, making clinical reasoning more accessible for resource-constrained deployments while still providing explainable decision support.