# PEA-DPO: Perception-Enhanced Alignment Direct Preference Optimization for MLLMs Alignment

arXiv: https://arxiv.org/abs/2608.19598 · HTML: https://mlsanity.com/paper/2608.19598/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Diffusion and multimodal model alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 28 of its 46 references that are in the map.
- First posted on arXiv 20 Aug 2026.

## Placed in

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

## Builds on (most cited first, 28 of 28)

- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [SimPO: Simple Preference Optimization with a Reference-Free Reward](https://mlsanity.com/paper/2405.14734/) (2024)
- [A Diagram Is Worth A Dozen Images](https://mlsanity.com/paper/1603.07396/) (2016)
- [Aligning Large Multimodal Models with Factually Augmented RLHF](https://mlsanity.com/paper/2309.14525/) (2023)
- [Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding](https://mlsanity.com/paper/2311.16922/) (2023)
- [Object Hallucination in Image Captioning](https://mlsanity.com/paper/1809.02156/) (2018)
- [OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation](https://mlsanity.com/paper/2311.17911/) (2023)
- [RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback](https://mlsanity.com/paper/2312.00849/) (2023)
- [AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluation](https://mlsanity.com/paper/2311.07397/) (2023)
- [Silkie: Preference Distillation for Large Visual Language Models](https://mlsanity.com/paper/2312.10665/) (2023)
- [Detecting and Mitigating Hallucination in Large Vision Language Models via Fine-Grained AI Feedback](https://mlsanity.com/paper/2404.14233/) (2024)
- [Enhancing Large Vision Language Models with Self-Training on Image Comprehension](https://mlsanity.com/paper/2405.19716/) (2024)
- [Mitigating Hallucinations in Large Vision-Language Models via DPO: On-Policy Data Hold the Key](https://mlsanity.com/paper/2501.09695/) (2025)
- [WPO: Enhancing RLHF with Weighted Preference Optimization](https://mlsanity.com/paper/2406.11827/) (2024)
- [CHiP: Cross-modal Hierarchical Direct Preference Optimization for Multimodal LLMs](https://mlsanity.com/paper/2501.16629/) (2025)
- [Enhancing Zero-Shot Vision Models by Label-Free Prompt Distribution Learning and Bias Correcting](https://mlsanity.com/paper/2410.19294/) (2024)
- [Understanding Alignment in Multimodal LLMs: A Comprehensive Study](https://mlsanity.com/paper/2407.02477/) (2024)
- [Enhancing CLIP Robustness via Cross-Modality Alignment](https://mlsanity.com/paper/2510.24038/) (2025)
- [Mitigating Hallucination Through Theory-Consistent Symmetric Multimodal Preference Optimization](https://mlsanity.com/paper/2506.11712/) (2025)
- [Res-Bench: Benchmarking the Robustness of Multimodal Large Language Models to Dynamic Resolution Input](https://mlsanity.com/paper/2510.16926/) (2025)
- [AdPO: Enhancing the Adversarial Robustness of Large Vision-Language Models with Preference Optimization](https://mlsanity.com/paper/2504.01735/) (2025)
- [Look Carefully: Adaptive Visual Reinforcements in Multimodal Large Language Models for Hallucination Mitigation](https://mlsanity.com/paper/2602.24041/) (2026)
- [LPOI: Listwise Preference Optimization for Vision Language Models](https://mlsanity.com/paper/2505.21061/) (2025)
- [Mitigating Hallucinations in Large Vision-Language Models without Performance Degradation](https://mlsanity.com/paper/2604.20366/) (2026)
- [Robustifying Vision-Language Models via Test-Time Prompt Adaptation](https://mlsanity.com/paper/2607.09450/) (2026)

## Abstract (arXiv)

Direct Preference Optimization (DPO) has emerged as an effective approach for aligning large language models (LLMs) with human preferences. However, its adaptation to multimodal settings remains unexplored. Through representational analysis, we identify a key limitation in multimodal preference optimization, which we term visual insensitivity: models often fail to distinguish between images and those with critical visual context removed. Our theoretical analysis further uncovers two manifestations of this problem, namely Across-Image Insensitivity and Within-Image Insensitivity. To address these challenges, we propose Perception-Enhanced Alignment DPO (PEA-DPO), a framework for multimodal LLMs alignment, which explicitly leverages visual preference signals to overcome visual insensitivity. We further provide a theoretical analysis demonstrating that PEA-DPO provably mitigates both failure modes. Empirical results demonstrate that PEA-DPO enhances sensitivity to visual context while preserving the language modeling capacity of the base model. Evaluations across three hallucination benchmarks using MLLMs of varying scales show that PEA-DPO effectively mitigates visual insensitivity, achieves stronger multimodal alignment, and substantially reduces hallucinations.