# Eeyore: Realistic Depression Simulation via Supervised and Preference Optimization

arXiv: https://arxiv.org/abs/2503.00018 · HTML: https://mlsanity.com/paper/2503.00018/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Medical LLM alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 12 of its 34 references that are in the map.
- First posted on arXiv 21 Feb 2025.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Medical LLM alignment](https://mlsanity.com/alignment/medical-llm-alignment/)

## Builds on (most cited first, 12 of 12)

- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Out of One, Many: Using Language Models to Simulate Human Samples](https://mlsanity.com/paper/2209.06899/) (2022)
- [Towards Emotional Support Dialog Systems](https://mlsanity.com/paper/2106.01144/) (2021)
- [Strengthening Multimodal Large Language Model with Bootstrapped Preference Optimization](https://mlsanity.com/paper/2403.08730/) (2024)
- [Roleplay-doh: Enabling Domain-Experts to Create LLM-simulated Patients via Eliciting and Adhering to Principles](https://mlsanity.com/paper/2407.00870/) (2024)
- [The Ethics of ChatGPT in Medicine and Healthcare: A Systematic Review on Large Language Models (LLMs)](https://mlsanity.com/paper/2403.14473/) (2024)
- [Interactive Agents: Simulating Counselor-Client Psychological Counseling via Role-Playing LLM-to-LLM Interactions](https://mlsanity.com/paper/2408.15787/) (2024)
- [Step-Controlled DPO: Leveraging Stepwise Error for Enhanced Mathematical Reasoning](https://mlsanity.com/paper/2407.00782/) (2024)
- [Speaker and Time-aware Joint Contextual Learning for Dialogue-act Classification in Counselling Conversations](https://mlsanity.com/paper/2111.06647/) (2021)
- [Towards a Client-Centered Assessment of LLM Therapists by Client Simulation](https://mlsanity.com/paper/2406.12266/) (2024)
- [Fine Tuning Large Language Models for Medicine: The Role and Importance of Direct Preference Optimization](https://mlsanity.com/paper/2409.12741/) (2024)

## Built on by (most cited first, 3 of 3)

- [MIRROR: Multimodal Cognitive Reframing Therapy for Rolling with Resistance](https://mlsanity.com/paper/2504.13211/) (2025)
- [Synthetic or Authentic? Building Mental Patient Simulators from Longitudinal Evidence](https://mlsanity.com/paper/2603.22704/) (2026)
- [Free Lunch for User Experience: Crowdsourcing Agents for Scalable User Studies](https://mlsanity.com/paper/2505.22981/) (2025)

## Abstract (arXiv)

Large Language Models (LLMs) have been previously explored for mental healthcare training and therapy client simulation, but they still fall short in authentically capturing diverse client traits and psychological conditions. We introduce Eeyore, an 8B model optimized for realistic depression simulation through a structured alignment framework, incorporating expert input at every stage. First, we systematically curate real-world depression-related conversations, extracting depressive traits to guide data filtering and psychological profile construction, and use this dataset to instruction-tune Eeyore for profile adherence. Next, to further enhance realism, Eeyore undergoes iterative preference optimization -- first leveraging model-generated preferences and then calibrating with a small set of expert-annotated preferences. Throughout the entire pipeline, we actively collaborate with domain experts, developing interactive interfaces to validate trait extraction and iteratively refine structured psychological profiles for clinically meaningful role-play customization. Despite its smaller model size, the Eeyore depression simulation outperforms GPT-4o with SOTA prompting strategies, both in linguistic authenticity and profile adherence.