ACPO: Counteracting Likelihood Displacement in Vision-Language Alignment with Asymmetric Constraints

• URL: http://arxiv.org/abs/2603.22165v1
• Date: Mon, 23 Mar 2026 16:26:11 GMT
• Title: ACPO: Counteracting Likelihood Displacement in Vision-Language Alignment with Asymmetric Constraints
• Abstract summary: We propose Asymmetric Constrained Preference Optimization (ACPO)<n>ACPO applies dynamic, target-oriented scaling to preference optimization.<n>We show that ACPO effectively reverses the chosen-reward degradation of standard DPO.
• Score: 26.309376587678354
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: While Direct Preference Optimization (DPO) has become the de facto approach for aligning Large Vision-Language Models (LVLMs), it suffers from Likelihood Displacement, where the probability of both chosen and rejected responses collapses. This optimization flaw is especially detrimental in multimodal settings: the erosion of chosen likelihoods -- a failure we term Visual Anchor Collapse -- causes models to abandon visual evidence for strong language priors, precipitating significant hallucinations. To address this, we propose Asymmetric Constrained Preference Optimization (ACPO), a modality-agnostic alignment mechanism that applies dynamic, target-oriented scaling to preference optimization. ACPO derives a complexity-aware scaling coefficient applied exclusively to the rejected reward, asymmetrically suppressing the gradient flow on the rejected term while preserving the chosen distribution as a gradient-stable reference. While fundamentally a general-purpose objective, breaking this gradient symmetry is crucial for multimodal tasks, as it mitigates the suppression of visual tokens by language priors. Experiments on InternVL models demonstrate that ACPO effectively reverses the chosen-reward degradation of standard DPO. By halting Visual Anchor Collapse, ACPO generally outperforms baselines on hallucination benchmarks (HallusionBench, MM-IFEval) and general leaderboards (MMBench, MMStar, OCRBenchV2) while driving concurrent improvements in general capabilities.

Related papers

• S-SPPO: Semantic-Calibrated Self-Play Preference Optimization [36.01916066772865]
  We introduce Self-Play Preference Optimization (SPPO), which iteratively refines the policy by training on self-generated win-lose pairs.<n>Our investigation reveals a critical instability in SPPO: the optimization is prone to policy degeneration when the preference assigns overly confident wins to semantically indistinguishable responses.<n>We show that S-SPPO avoids the performance degradation seen in prior methods, achieving 52.19% win rate and 47.46% length-controlled win rate on AlpacaEval 2.0 with Llama-3-8B, without using additional human-annotated preferences during training.
  arXiv  Detail & Related papers  (2026-06-01T02:06:58Z)
• Gradient-Gated DPO: Stabilizing Preference Optimization in Language Models [0.0]
  We introduce Gradient-Gated Preference Optimization (Gate-DPO)<n>Gate-DPO stabilizes training by modulating rejected gradients according to the model's probability geometry.<n>Experiments across multiple architectures and preference datasets show that Gate-DPO consistently reduces squeezing and improves chosen-response likelihood.
  arXiv  Detail & Related papers  (2026-05-04T14:15:24Z)
• Hindsight-Anchored Policy Optimization: Turning Failure into Feedback in Sparse Reward Settings [7.554296968064274]
  We present Hindsight-Anchored Policy optimization (HAPO) for Reinforcement Learning with Verifiable Rewards (RLVR)<n>HAPO employs a hindsight mechanism that selectively anchors optimization to teacher demonstrations during failure.<n>We show that HAPO achieves textitasymptotic consistency: by naturally annealing the teacher signal as the policy improves, HAPO recovers the unbiased on-policy gradient.
  arXiv  Detail & Related papers  (2026-03-11T21:33:41Z)
• Anchored Policy Optimization: Mitigating Exploration Collapse Via Support-Constrained Rectification [14.911955979675772]
  We propose Anchored Policy Optimization (APO) to shift the paradigm from global Shape Matching to Support Coverage.<n>APO breaks the accuracy-diversity trade-off, significantly improving Pass@1 while restoring the Pass@K diversity typically lost by standard policy gradient methods.
  arXiv  Detail & Related papers  (2026-02-05T14:41:57Z)
• MAESTRO: Meta-learning Adaptive Estimation of Scalarization Trade-offs for Reward Optimization [56.074760766965085]
  Group-Relative Policy Optimization has emerged as an efficient paradigm for aligning Large Language Models (LLMs)<n>We propose MAESTRO, which treats reward scalarization as a dynamic latent policy, leveraging the model's terminal hidden states as a semantic bottleneck.<n>We formulate this as a contextual bandit problem within a bi-level optimization framework, where a lightweight Conductor network co-evolves with the policy by utilizing group-relative advantages as a meta-reward signal.
  arXiv  Detail & Related papers  (2026-01-12T05:02:48Z)
• Anchoring Values in Temporal and Group Dimensions for Flow Matching Model Alignment [61.80228667422234]
  VGPO redefines value estimation across both temporal and group dimensions.<n>It transforms the sparse terminal reward into dense, process-aware value estimates.<n>It replaces standard group normalization with a novel process enhanced by absolute values to maintain a stable optimization signal.
  arXiv  Detail & Related papers  (2025-12-13T16:31:26Z)
• AMaPO: Adaptive Margin-attached Preference Optimization for Language Model Alignment [25.526336903358757]
  offline preference optimization offers a simpler and more stable alternative to RLHF for aligning language models.<n>We propose Adaptive Margin-attached Preference Optimization (AMaPO), a simple yet principled algorithm.<n>AMaPO employs an instance-wise adaptive margin, refined by Z-normalization and exponential scaling, which dynamically reallocates learning effort by amplifying gradients for misranked samples and suppressing them for correct ones.
  arXiv  Detail & Related papers  (2025-11-12T14:51:59Z)
• Mitigating Hallucination Through Theory-Consistent Symmetric Multimodal Preference Optimization [69.05600758833471]
  Direct Preference Optimization (DPO) has emerged as an effective approach for mitigating hallucination in Multimodal Large Language Models (MLLMs)<n>We propose a Symmetric Multimodal Preference Optimization (SymMPO) which conducts symmetric preference learning with direct preference supervision (i.e., response pairs)<n>In addition to conventional ordinal preference learning, SymMPO introduces a preference margin consistency loss to quantitatively regulate the preference gap between symmetric preference pairs.
  arXiv  Detail & Related papers  (2025-06-13T12:29:15Z)
• Proximalized Preference Optimization for Diverse Feedback Types: A Decomposed Perspective on DPO [19.5712961932773]
  We revisit direct preference optimization (DPO) and demonstrate that its loss theoretically admits a decomposed reformulation.<n>We introduce PRoximalized PReference Optimization (PRO), a unified method to align with diverse feeback types.
  arXiv  Detail & Related papers  (2025-05-29T10:23:22Z)
• 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)
• 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)
• Correcting the Mythos of KL-Regularization: Direct Alignment without Overoptimization via Chi-Squared Preference Optimization [78.82586283794886]
  $chi2$-Preference Optimization ($chi$PO) is an efficient offline alignment algorithm provably robust to overoptimization.<n>$chi$PO implements the principle of pessimism in the face of uncertainty via regularization.<n>$chi$PO's simplicity and strong guarantees make it the first practical and general-purpose offline alignment algorithm provably robust to overoptimization.
  arXiv  Detail & Related papers  (2024-07-18T11:08:40Z)
• 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)

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.