# Preference Robustness for DPO with Applications to Public Health

arXiv: https://arxiv.org/abs/2509.02709 · HTML: https://mlsanity.com/paper/2509.02709/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in LLM preference optimization methods, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 8 of its 45 references that are in the map.
- First posted on arXiv 2 Sep 2025.

## Placed in

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

## Builds on (most cited first, 8 of 8)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for Problem-Solving with Language Models](https://mlsanity.com/paper/2408.00724/) (2024)
- [A Comprehensive Survey of Direct Preference Optimization: Datasets, Theories, Variants, and Applications](https://mlsanity.com/paper/2410.15595/) (2024)
- [Robust LLM Alignment via Distributionally Robust Direct Preference Optimization](https://mlsanity.com/paper/2502.01930/) (2025)
- [Distributionally Robust Reinforcement Learning with Human Feedback](https://mlsanity.com/paper/2503.00539/) (2025)
- [Balancing Act: Prioritization Strategies for LLM-Designed Restless Bandit Rewards](https://mlsanity.com/paper/2408.12112/) (2024)
- [Orthogonal Finetuning for Direct Preference Optimization](https://mlsanity.com/paper/2409.14836/) (2024)
- [Towards Understanding Why Data Augmentation Improves Generalization](https://mlsanity.com/paper/2502.08940/) (2025)

## Abstract (arXiv)

We study an LLM fine-tuning task for designing reward functions for sequential resource allocation problems in public health, guided by human preferences expressed in natural language. This setting presents a challenging testbed for alignment due to complex and ambiguous objectives and limited data availability. We propose DPO-PRO, a robust fine-tuning algorithm based on Direct Preference Optimization (DPO), which accounts for uncertainty in the preference distribution using a lightweight Distributionally Robust Optimization (DRO) formulation. Unlike prior DRO-based DPO methods, DPO-PRO is significantly less conservative. We evaluate DPO-PRO on a real-world maternal mobile health program operated by the non-profit organization ARMMAN, as well as on standard alignment benchmarks. Experimental results demonstrate that our method consistently improves robustness to noisy preference signals compared to existing DPO variants. Moreover, DPO-PRO achieves comparable performance to prior self-reflection-based baseline for reward function design, while requiring significantly lower inference-time cost.