# Aligning Diffusion Language Models via Unpaired Preference Optimization

arXiv: https://arxiv.org/abs/2510.23658 · HTML: https://mlsanity.com/paper/2510.23658/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models.
- Builds on 25 of its 28 references that are in the map.
- First posted on arXiv 26 Oct 2025.

## Placed in

- Diffusion models / [Diffusion and flow policies for robotics](https://mlsanity.com/diffusion/diffusion-and-flow-policies-for-robotics/)
- LLM alignment / RLHF / preference optimisation / [Diffusion and multimodal model alignment](https://mlsanity.com/alignment/diffusion-and-multimodal-model-alignment/)

## Builds on (most cited first, 25 of 25)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [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)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Structured Denoising Diffusion Models in Discrete State-Spaces](https://mlsanity.com/paper/2107.03006/) (2021)
- [SimPO: Simple Preference Optimization with a Reference-Free Reward](https://mlsanity.com/paper/2405.14734/) (2024)
- [A General Theoretical Paradigm to Understand Learning from Human Preferences](https://mlsanity.com/paper/2310.12036/) (2023)
- [Large Language Diffusion Models](https://mlsanity.com/paper/2502.09992/) (2025)
- [Simple and Effective Masked Diffusion Language Models](https://mlsanity.com/paper/2406.07524/) (2024)
- [Diffusion Model Alignment Using Direct Preference Optimization](https://mlsanity.com/paper/2311.12908/) (2023)
- [Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution](https://mlsanity.com/paper/2310.16834/) (2023)
- [Dream 7B: Diffusion Large Language Models](https://mlsanity.com/paper/2508.15487/) (2025)
- [RRHF: Rank Responses to Align Language Models with Human Feedback without tears](https://mlsanity.com/paper/2304.05302/) (2023)
- [Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Data](https://mlsanity.com/paper/2406.03736/) (2024)
- [LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Models](https://mlsanity.com/paper/2505.19223/) (2025)
- [Scaling Diffusion Language Models via Adaptation from Autoregressive Models](https://mlsanity.com/paper/2410.17891/) (2024)
- [MMaDA: Multimodal Large Diffusion Language Models](https://mlsanity.com/paper/2505.15809/) (2025)
- [Mercury: Ultra-Fast Language Models Based on Diffusion](https://mlsanity.com/paper/2506.17298/) (2025)
- [d1: Scaling Reasoning in Diffusion Large Language Models via Reinforcement Learning](https://mlsanity.com/paper/2504.12216/) (2025)
- [Aligning Diffusion Models by Optimizing Human Utility](https://mlsanity.com/paper/2404.04465/) (2024)
- [wd1: Weighted Policy Optimization for Reasoning in Diffusion Language Models](https://mlsanity.com/paper/2507.08838/) (2025)
- [Margin-aware Preference Optimization for Aligning Diffusion Models without Reference](https://mlsanity.com/paper/2406.06424/) (2024)
- [Calibrated Multi-Preference Optimization for Aligning Diffusion Models](https://mlsanity.com/paper/2502.02588/) (2025)
- [Discrete Diffusion Trajectory Alignment via Stepwise Decomposition](https://mlsanity.com/paper/2507.04832/) (2025)

## Built on by (most cited first, 1 of 1)

- [Backdooring Masked Diffusion Language Models](https://mlsanity.com/paper/2605.19262/) (2026)

## Abstract (arXiv)

Diffusion language models (dLLMs) are an emerging alternative to autoregressive (AR) generators, but aligning them to human preferences is challenging because sequence log-likelihoods are intractable and pairwise preference data are costly to collect. We introduce ELBO-KTO, which combines an ELBO surrogate for diffusion log-likelihoods with a prospect-theoretic, unpaired preference objective (Kahneman Tversky Optimization, KTO). We analyze the bias and variance induced by the ELBO substitution and employ variance-reduction practices that stabilize gradients during training. Applied to LLaDA-8B-Instruct, ELBO-KTO yields 65.9% and 62.3% adjusted win rates on kto-mix-14k and UltraFeedback-Binary, respectively, versus the base model under an automatic LLM judge. Across downstream tasks, including GSM8K, MMLU, and additional reasoning/knowledge benchmarks, ELBO-KTO trained on UltraFeedback-Binary performs on par with or better than the base model under identical decoding. This establishes unpaired preference optimization as a viable alternative to pairwise alignment in diffusion LLMs.