Why DPO is a Misspecified Estimator and How to Fix It

• URL: http://arxiv.org/abs/2510.20413v1
• Date: Thu, 23 Oct 2025 10:30:29 GMT
• Title: Why DPO is a Misspecified Estimator and How to Fix It
• Authors: Aditya Gopalan, Sayak Ray Chowdhury, Debangshu Banerjee,
• Abstract summary: We show that DPO encodes a statistical estimation problem over reward functions induced by a parametric policy class.<n>We propose AuxDPO, which introduces additional auxiliary variables in the DPO loss function to help move towards the RLHF solution.
• Score: 17.3051205548111
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Direct alignment algorithms such as Direct Preference Optimization (DPO) fine-tune models based on preference data, using only supervised learning instead of two-stage reinforcement learning with human feedback (RLHF). We show that DPO encodes a statistical estimation problem over reward functions induced by a parametric policy class. When the true reward function that generates preferences cannot be realized via the policy class, DPO becomes misspecified, resulting in failure modes such as preference order reversal, worsening of policy reward, and high sensitivity to the input preference data distribution. On the other hand, we study the local behavior of two-stage RLHF for a parametric class and relate it to a natural gradient step in policy space. Our fine-grained geometric characterization allows us to propose AuxDPO, which introduces additional auxiliary variables in the DPO loss function to help move towards the RLHF solution in a principled manner and mitigate the misspecification in DPO. We empirically demonstrate the superior performance of AuxDPO on didactic bandit settings as well as LLM alignment tasks.

Related papers

• Preference Robustness for DPO with Applications to Public Health [26.99327564250612]
  We propose DPO-PRO, a robust fine-tuning algorithm based on Direct Preference Optimization (DPO)<n>We evaluate DPO-PRO on a real-world maternal mobile health program operated by the non-profit organization ARMMAN.
  arXiv  Detail & Related papers  (2025-09-02T18:10:32Z)
• Understanding the Performance Gap in Preference Learning: A Dichotomy of RLHF and DPO [53.914481648817066]
  We decompose the performance gap between reinforcement learning from human feedback and direct preference optimization under a representation gap.<n>We show that RLHF, DPO, or online DPO can outperform one another depending on type of model mis-specifications.
  arXiv  Detail & Related papers  (2025-05-26T09:54:02Z)
• C2-DPO: Constrained Controlled Direct Preference Optimization [22.730518243326394]
  Direct preference optimization (textttDPO) has emerged as a promising approach for solving the alignment problem in AI.<n>We show that textttDPO loss could be derived by starting from an alternative optimization problem that only defines the KL guardrail on in-sample responses.
  arXiv  Detail & Related papers  (2025-02-22T00:38:44Z)
• Entropy Controllable Direct Preference Optimization [3.536605202672355]
  We propose a simple modification to DPO, H-DPO, which allows for control over the entropy of the resulting policy.<n>In our experiments, we show that H-DPO outperformed DPO across various tasks, demonstrating superior results in pass@$k$ evaluations for mathematical tasks.
  arXiv  Detail & Related papers  (2024-11-12T07:09:44Z)
• Uncertainty-Penalized Direct Preference Optimization [52.387088396044206]
  We develop a pessimistic framework for DPO by introducing preference uncertainty penalization schemes. The penalization serves as a correction to the loss which attenuates the loss gradient for uncertain samples. We show improved overall performance compared to vanilla DPO, as well as better completions on prompts from high-uncertainty chosen/rejected responses.
  arXiv  Detail & Related papers  (2024-10-26T14:24:37Z)
• AlphaDPO: Adaptive Reward Margin for Direct Preference Optimization [45.46582930202524]
  $alpha$-DPO is an adaptive preference optimization algorithm for large language models.<n>It balances the policy model and the reference model to achieve personalized reward margins.<n>It consistently outperforms DPO and SimPO across various model settings.
  arXiv  Detail & Related papers  (2024-10-14T04:29:57Z)
• Zeroth-Order Policy Gradient for Reinforcement Learning from Human Feedback without Reward Inference [15.038210624870656]
  Reward inference is a critical intermediate step in the Reinforcement Learning from Human Feedback pipeline.<n>This paper develops two RLHF algorithms without reward inference for general RL problems beyond bandits and deterministic MDP bandit, and general preference models beyond the Bradley-Terry model.
  arXiv  Detail & Related papers  (2024-09-25T22:20:11Z)
• Provably Mitigating Overoptimization in RLHF: Your SFT Loss is Implicitly an Adversarial Regularizer [52.09480867526656]
  We identify the source of misalignment as a form of distributional shift and uncertainty in learning human preferences.<n>To mitigate overoptimization, we first propose a theoretical algorithm that chooses the best policy for an adversarially chosen reward model.<n>Using the equivalence between reward models and the corresponding optimal policy, the algorithm features a simple objective that combines a preference optimization loss and a supervised learning loss.
  arXiv  Detail & Related papers  (2024-05-26T05:38:50Z)
• D2PO: Discriminator-Guided DPO with Response Evaluation Models [63.71853401569461]
  We propose D2PO, discriminator-guided DPO, for the online setting where preferences are being collected throughout learning. As we collect gold preferences, we use these not only to train our policy, but to train a discriminative response evaluation model to silver-label even more synthetic data for policy training. We show conditions under which silver labeling is most helpful: it is most effective when training the policy with DPO, outperforming traditional PPO, and benefits from maintaining a separate discriminator from the policy model.
  arXiv  Detail & Related papers  (2024-05-02T17:44:41Z)
• From $r$ to $Q^*$: Your Language Model is Secretly a Q-Function [50.812404038684505]
  We show that we can derive DPO in the token-level MDP as a general inverse Q-learning algorithm, which satisfies the Bellman equation. We discuss applications of our work, including information elicitation in multi-turn dialogue, reasoning, agentic applications and end-to-end training of multi-model systems.
  arXiv  Detail & Related papers  (2024-04-18T17:37:02Z)
• Learn Your Reference Model for Real Good Alignment [3.091688550418396]
  offline methods for Large Language Models (LLMs) alignment are susceptible to overoptimization.<n>We propose a new paradigm of offline alignment methods, called Trust Region, which dynamically updates the reference policy throughout the training process.<n>Our results show that TR alignment methods effectively mitigate overoptimization, enabling models to maintain strong performance even when substantially deviating from the initial reference policy.
  arXiv  Detail & Related papers  (2024-04-15T10:44:31Z)
• Human Alignment of Large Language Models through Online Preference Optimisation [50.52545798589968]
  We show the equivalence between two recent alignment methods, namely Identity Policy optimisation (IPO) and Nash Mirror Descent (Nash-MD) This equivalence can be proven when we consider the online version of IPO, that is when both generations are sampled by the online policy and annotated by a trained preference model. We introduce the IPO-MD algorithm that generates data with a mixture policy (between the online and reference policy) similarly as the general Nash-MD algorithm.
  arXiv  Detail & Related papers  (2024-03-13T15:47:26Z)
• RS-DPO: A Hybrid Rejection Sampling and Direct Preference Optimization Method for Alignment of Large Language Models [7.676477609461592]
  Reinforcement learning from human feedback (RLHF) has been extensively employed to align large language models with user intent. DPO relies on contrastive responses generated from human annotator and alternative LLM, instead of the policy model. In this paper, we address both challenges by systematically combining sampling rejection (RS) and DPO. Our proposed method effectively fine-tunes LLMs with limited resource environments, leading to improved alignment with user intent.
  arXiv  Detail & Related papers  (2024-02-15T16:00:58Z)

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.