# Aligning Medical Conversational AI through Online Reinforcement Learning with Information-Theoretic Rewards

arXiv: https://arxiv.org/abs/2601.17828 · HTML: https://mlsanity.com/paper/2601.17828/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 47 references that are in the map.
- First posted on arXiv 25 Jan 2026.

## Builds on (most cited first, 16 of 16)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models](https://mlsanity.com/paper/2401.01335/) (2024)
- [Reinforced Self-Training (ReST) for Language Modeling](https://mlsanity.com/paper/2308.08998/) (2023)
- [ChatDoctor: A Medical Chat Model Fine-Tuned on a Large Language Model Meta-AI (LLaMA) Using Medical Domain Knowledge](https://mlsanity.com/paper/2303.14070/) (2023)
- [BioMistral: A Collection of Open-Source Pretrained Large Language Models for Medical Domains](https://mlsanity.com/paper/2402.10373/) (2024)
- [MedAlpaca -- An Open-Source Collection of Medical Conversational AI Models and Training Data](https://mlsanity.com/paper/2304.08247/) (2023)
- [Deal or No Deal? End-to-End Learning for Negotiation Dialogues](https://mlsanity.com/paper/1706.05125/) (2017)
- [HuaTuo: Tuning LLaMA Model with Chinese Medical Knowledge](https://mlsanity.com/paper/2304.06975/) (2023)
- [MedDialog: Two Large-scale Medical Dialogue Datasets](https://mlsanity.com/paper/2004.03329/) (2020)
- [ClinicalGPT: Large Language Models Finetuned with Diverse Medical Data and Comprehensive Evaluation](https://mlsanity.com/paper/2306.09968/) (2023)
- [End-to-End Knowledge-Routed Relational Dialogue System for Automatic Diagnosis](https://mlsanity.com/paper/1901.10623/) (2019)
- [Information Maximizing Visual Question Generation](https://mlsanity.com/paper/1903.11207/) (2019)
- [Real-World Doctor Agent with Proactive Consultation through Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2505.19630/) (2025)
- [Reflect-RL: Two-Player Online RL Fine-Tuning for LMs](https://mlsanity.com/paper/2402.12621/) (2024)
- [Answerer in Questioner's Mind: Information Theoretic Approach to Goal-Oriented Visual Dialog](https://mlsanity.com/paper/1802.03881/) (2018)

## Abstract (arXiv)

We present Information Gain Fine-Tuning (IGFT), a novel approach for training medical conversational AI to conduct effective patient interviews and generate comprehensive History of Present Illness (HPI) without requiring pre-collected human conversations. IGFT combines online Group Relative Policy Optimization (GRPO) with information-theoretic rewards, enabling models to learn from self-generated conversations with simulated patients. Unlike existing approaches that rely on expensive expert-annotated conversations or static datasets, our online RL framework allows models to discover effective questioning strategies through exploration. Our key innovation is an information gain reward function that tracks which clinical entities such as symptoms, temporal patterns, and medical history, are revealed during conversation. Each question's reward is computed based on its expected information gain combined with GPT-4o-mini quality assessments across dimensions including clinical relevance, patient engagement, and specificity. This hybrid approach ensures models learn to ask targeted, clinically appropriate questions that efficiently gather diagnostic information. We fine-tune two models using LoRA: Llama-3.1-8B-Instruct and DeepSeek-R1-Distill-Qwen-7B (a reasoning-optimized model). Training exclusively on Avey data containing concise HPIs, we evaluate generalization to MIMIC data with longer, more elaborate HPIs. DeepSeek-R1-Distill-Qwen-7B (IGFT) achieves F1 scores of 0.408 on Avey (10.9% improvement over base) and 0.289 on MIMIC (12.9% improvement), while Llama-3.1-8B-Instruct (IGFT) reaches 0.384 and 0.336 respectively. Both models outperform OpenAI's model on MIMIC and surpass medical domain-specific baselines like HuatuoGPT and UltraMedical, which were optimized for single-turn medical QA rather than multi-turn conversations.