What Is Preference Optimization Doing, How and Why?

• URL: http://arxiv.org/abs/2512.00778v1
• Date: Sun, 30 Nov 2025 08:27:59 GMT
• Title: What Is Preference Optimization Doing, How and Why?
• Authors: Yue Wang, Qizhou Wang, Zizhuo Zhang, Ang Li, Gang Niu, Bo Han, Masashi Sugiyama,
• Abstract summary: A common belief is that DPO is supervised learning while PPO is reinforcement learning, yet deeper analyses for the reasons underlying these differences remain lacking.<n>First, we examine the target directions of gradient-based updates and find that DPO follows stable targets, whereas PPO follows dynamic targets that balance exploration and exploitation.<n>Second, we examine the roles of positive learning, negative learning, and loss reweighting, which are three key components in PO methods.
• Score: 73.78865043839939
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Preference optimization (PO) is indispensable for large language models (LLMs), with methods such as direct preference optimization (DPO) and proximal policy optimization (PPO) achieving great success. A common belief is that DPO is supervised learning while PPO is reinforcement learning, yet deeper analyses for the reasons underlying these differences remain lacking. To fill this gap, we analyze their optimization dynamics, revealing distinct algorithmic behaviors and comprehending their underlying causes. First, we examine the target directions of gradient-based updates and find that DPO follows stable targets, whereas PPO follows dynamic targets that balance exploration and exploitation, thus validating the common belief from a new perspective. Second, we examine the roles of positive learning, negative learning, and loss reweighting, which are three key components in PO methods. Our analyses reveal that these components play fairly different roles. In DPO, positive and negative learning jointly shape the learning targets meanwhile mutually offset each other. However, loss reweighting in DPO acts less as a reward signal but more as a regularizer to mitigate overfitting. In PPO, negative learning primarily supports exploration rather than determining the targets. Meanwhile, loss reweighting, related to absolute values of token-level advantages, indicates the distinct roles of token groups in updating targets. Given these findings, we conduct carefully designed ablation studies to further examine how controlling these dynamics impacts optimization efficiency and practical performance. The insights gained from our analyses not only deepen the understanding of PO methods but also inspire the development of more preference-aligned LLMs.

Related papers

• $φ$-DPO: Fairness Direct Preference Optimization Approach to Continual Learning in Large Multimodal Models [58.217707070069885]
  This paper presents a novel Fairness Direct Preference Optimization (FaiDPO or $$-DPO) framework for continual learning in LMMs.<n>We first propose a new continual learning paradigm based on Direct Preference Optimization (DPO) to mitigate catastrophic forgetting by aligning learning with pairwise preference signals.<n> Extensive experiments and ablation studies show the proposed $$-DPO achieves State-of-the-Art performance across multiple benchmarks.
  arXiv  Detail & Related papers  (2026-02-26T04:14:33Z)
• BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping [69.74252624161652]
  We propose BAlanced Policy Optimization with Adaptive Clipping (BAPO)<n>BAPO dynamically adjusts clipping bounds to adaptively re-balance positive and negative contributions, preserve entropy, and stabilize RL optimization.<n>On AIME 2024 and AIME 2025 benchmarks, our 7B BAPO model surpasses open-source counterparts such as SkyWork-OR1-7B.
  arXiv  Detail & Related papers  (2025-10-21T12:55:04Z)
• On-Policy RL with Optimal Reward Baseline [109.47676554514193]
  On-Policy RL with Optimal reward baseline (OPO) is a novel and simplified reinforcement learning algorithm.<n>OPO emphasizes the importance of exact on-policy training, which empirically stabilizes the training process and enhances exploration.<n>Results demonstrate OPO's superior performance and training stability without additional models or regularization terms.
  arXiv  Detail & Related papers  (2025-05-29T15:58:04Z)
• On the Effect of Negative Gradient in Group Relative Deep Reinforcement Optimization [52.76330545825083]
  Reinforcement learning (RL) has become popular in enhancing the reasoning capabilities of large language models (LLMs)<n>We identify a previously unrecognized phenomenon we term Lazy Likelihood Displacement (LLD), wherein the likelihood of correct responses marginally increases or even decreases during training.<n>We develop a method called NTHR, which downweights penalties on tokens contributing to the LLD. Unlike prior DPO-based approaches, NTHR takes advantage of GRPO's group-based structure, using correct responses as anchors to identify influential tokens.
  arXiv  Detail & Related papers  (2025-05-24T18:58:51Z)
• Gradient Imbalance in Direct Preference Optimization [26.964127989679596]
  We propose Balanced-DPO, a simple yet effective modification to the DPO objective that introduces a computationally efficient gradient reweighting mechanism.<n>Our experiments demonstrate the effectiveness of Balanced-DPO, validating the theoretical findings and confirming that addressing gradient imbalance is key to improving DPO's performance.
  arXiv  Detail & Related papers  (2025-02-28T08:47:03Z)
• TIS-DPO: Token-level Importance Sampling for Direct Preference Optimization With Estimated Weights [73.9088920210495]
  We propose a token-level importance sampling DPO objective named TIS-DPO that assigns importance weights to each token based on its reward.<n>TIS-DPO significantly outperforms various baseline methods on harmlessness and helpfulness alignment and summarization tasks.
  arXiv  Detail & Related papers  (2024-10-06T04:03:00Z)
• 3D-Properties: Identifying Challenges in DPO and Charting a Path Forward [17.27880657597116]
  We revisit DPO, analyzing its theoretical foundations and empirical performance.<n>We identify three key properties, termed 3D properties, that emerge from DPO's learning process.<n>We propose simple regularization techniques that improve training stability and performance.
  arXiv  Detail & Related papers  (2024-06-11T14:59:24Z)
• Direct Alignment of Language Models via Quality-Aware Self-Refinement [31.845241241178982]
  We investigate the use of intrinsic knowledge within the on-the-fly fine-tuning LLM to obtain relative qualities and help to refine the loss function. We show that the constructed refinement function can help self-refine the loss function under mild assumptions. Experiments indicate that they can improve the performance of the fine-tuned models over DPO and IPO.
  arXiv  Detail & Related papers  (2024-05-31T17:31:18Z)
• Towards Analyzing and Understanding the Limitations of DPO: A Theoretical Perspective [25.34250859820326]
  We provide an analytical framework using the field theory to analyze the optimization process of DPO. We find that the DPO loss function decreases the probability of producing human dispreferred data at a faster rate than it increases the probability of producing preferred data.
  arXiv  Detail & Related papers  (2024-04-06T13:24:37Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.