Consistency as Inductive Bias: Learning Cross-View Invariance for Robust Multimodal Reasoning

• URL: http://arxiv.org/abs/2606.29812v1
• Date: Mon, 29 Jun 2026 05:45:18 GMT
• Title: Consistency as Inductive Bias: Learning Cross-View Invariance for Robust Multimodal Reasoning
• Abstract summary: We propose ConsistRoll to inject cross-view consistency into RLVR training.<n>It places original and semantically invariant transformed views in the same generation group.<n>It assigns a joint reward only when paired completions are both correct and consistent.
• Score: 40.68014726208686
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Inductive biases steer learning toward generalizable solutions by encoding task structure. In this work, we identify a crucial missing bias in MLLMs: cross-view consistency, \textit{i.e.}, semantically invariant views of the same instance should lead to the same answer. Standard reinforcement learning with verifiable rewards (RLVR) objectives do not impose this constraint, but instead assign pointwise rewards to each visual input. Even with data augmentation (DA), transformed views are typically rewarded independently, providing little signal once within-view rewards saturate. We propose \textbf{ConsistRoll}, a simple but effective method that injects cross-view consistency into RLVR training by reusing the group-sampling mechanism of GRPO. Specifically, ConsistRoll places original and semantically invariant transformed views in the same generation group, and assigns a joint reward only when paired completions are both correct and consistent. In this way, ConsistRoll turns consistency into an online credit-assignment signal, \textbf{without extra generation overhead and annotations}. Theoretically, we show that cross-view consistency is a valid inductive bias, and ConsistRoll introduces a cross-view correction term absent from DA, penalizing view dependence and alleviating advantage collapse. Comprehensive benchmarks across math, general-purpose, hallucination domains confirm that ConsistRoll achieves robust improvements in multimodal reasoning.

Related papers

• Unpaired Modality-Agnostic Generative Recommendation [26.342219762853187]
  UnpairGR learns a unified semantic-ID space from paired, image-only, and text-only observations.<n>Experiments on three benchmark datasets demonstrate that UnpairGR consistently improves recommendation performance.
  arXiv  Detail & Related papers  (2026-08-03T16:46:47Z)
• SIVA-RL: Sensitivity-Invariance Visual Alignment for Multimodal Reinforcement Learning [19.079142995255506]
  We propose SIVA-RL to replace operator-conditioned regularization with sample-wise, outcome-conditioned supervision.<n>SIVA-RL constructs localized interventions through token-aligned, distance-constrained within-image Patch.<n>It yields an 8.79 percentage-point gain on vision-dependent reasoning and up to 14.9% relative overall improvement across all four GRPO- and DAPO-based configurations.
  arXiv  Detail & Related papers  (2026-07-15T15:13:02Z)
• C$^{2}$R: Cross-sample Consistency Regularization Mitigates Feature Splitting and Absorption in Sparse Autoencoders [56.67585330425431]
  Sparse Autoencoders (SAEs) are widely used to interpret large language models by decomposing activations into sparse, human-understandable features.<n>These issues stem from inconsistent latent assignment across samples.<n>We introduce C$2$R, which penalizes the co-activation of directionally similar latents.
  arXiv  Detail & Related papers  (2026-06-29T17:45:31Z)
• TRACE: A Unified Rollout Budget Allocation Framework for Efficient Agentic Reinforcement Learning [52.67756371231985]
  Reinforcement learning with verifiable rewards (RLVR) is a promising approach for enhancing reasoning and agentic behavior in large language models.<n>We introduce Tree Rollout Allocation for Contrastive Exploration (TRACE), a unified rollout allocation framework.<n>Technically, TRACE allocates rollout budget to both prompt roots and intermediate prefixes that are most likely to yield mixed terminal rewards.
  arXiv  Detail & Related papers  (2026-06-09T17:16:03Z)
• DIVA: Harnessing the Representation Divergence in Unified Multimodal Models for Mutual Reinforcement [42.87874090062771]
  Unified Multimodal models (UMMs) have shown impressive performance in both understanding and generation.<n>We identify a fundamental challenge that lies in inductive biases induced by distinct supervision signals.<n>We propose DIVA, a self-improved post-training framework that transforms the representation divergence into interior synergy.
  arXiv  Detail & Related papers  (2026-05-25T01:17:32Z)
• LC-ERD: Mining Latent Logic for Self-Evolving Reasoning via Consistency-Regulated Reward Decomposition [55.572260012037084]
  We introduce LC-ERD (Logic-Consistent Endogenous Reward Decomposition), a framework framing self-alignment as latent structure mining.<n>We derive a Variational Logic Potential by aggregating consensus from the model's Latent Logic Expertise.<n>Experiments show LC-ERD delivers a robust self-evolution path, uncovering trade-offs between logic consistency and accuracy.
  arXiv  Detail & Related papers  (2026-05-19T07:27:50Z)
• Stabilizing Unsupervised Self-Evolution of MLLMs via Continuous Softened Retracing reSampling [19.22335478070254]
  We propose Continuous Softened Retracing reSampling (CSRS) in MLLM self-evolution.<n>Specifically, we introduce a Retracing Re-inference Mechanism (RRM) that the model re-inferences from anchor points to expand the exploration of long-tail reasoning paths.<n>We demonstrate that CSRS significantly enhances the reasoning performance of Qwen2.5-VL-7B on benchmarks such as MathVision.
  arXiv  Detail & Related papers  (2026-04-04T08:52:43Z)
• Soft Equivariance Regularization for Invariant Self-Supervised Learning [23.047550451521662]
  Self-supervised learning (SSL) typically learns representations invariant to semantic-preserving augmentations.<n>We propose Soft Equivariance Regularization (SER), a plug-in regularizer that decouples where invariance and equivariance are enforced.<n>SER learns/predicts no per-sample transformation codes/labels, requires no auxiliary transformation-prediction head, and adds only 1.008x training FLOPs.
  arXiv  Detail & Related papers  (2026-03-04T13:36:17Z)
• From Verifiable Dot to Reward Chain: Harnessing Verifiable Reference-based Rewards for Reinforcement Learning of Open-ended Generation [52.62655622099456]
  We propose reinforcement learning with verifiable reference-based rewards (RLVRR)<n>Instead of checking the final answer, RLVRR extracts an ordered linguistic signal from high-quality references (i.e., reward chain)<n>In this way, RLVRR decomposes rewards into two dimensions: content, which preserves deterministic core concepts, and style, which evaluates adherence to stylistic properties.
  arXiv  Detail & Related papers  (2026-01-26T14:39:58Z)
• CARE What Fails: Contrastive Anchored-REflection for Verifiable Multimodal [84.71254539482369]
  Group-relative reinforcement learning with verifiable rewards (RLVR) often wastes the most informative data it already has the failures.<n>We present CARE, a failure-centric post-training framework for multimodal reasoning that turns errors into supervision.<n> CARE improves accuracy and training smoothness while explicitly increasing the share of learning signal that comes from failures.
  arXiv  Detail & Related papers  (2025-12-22T16:34:21Z)
• Hybrid Reinforcement: When Reward Is Sparse, It's Better to Be Dense [36.71358559780692]
  HERO is a reinforcement learning framework that integrates verifier signals with reward-model scores in a structured way.<n> HERO consistently outperforms RM-only and verifier-only baselines, with strong gains on both verifiable and hard-to-verify tasks.
  arXiv  Detail & Related papers  (2025-10-08T17:09:41Z)
• Cross-modal Full-mode Fine-grained Alignment for Text-to-Image Person Retrieval [54.90229711181207]
  Text-to-Image Person Retrieval (TIPR) aims to retrieve the most relevant person images based on a given text query.<n>The key challenge in TIPR lies in achieving effective alignment between textual and visual modalities.<n>We propose FMFA, a cross-modal Full-Mode Fine-grained Alignment framework.
  arXiv  Detail & Related papers  (2025-09-17T07:12:05Z)
• Relieving Long-tailed Instance Segmentation via Pairwise Class Balance [85.53585498649252]
  Long-tailed instance segmentation is a challenging task due to the extreme imbalance of training samples among classes. It causes severe biases of the head classes (with majority samples) against the tailed ones. We propose a novel Pairwise Class Balance (PCB) method, built upon a confusion matrix which is updated during training to accumulate the ongoing prediction preferences.
  arXiv  Detail & Related papers  (2022-01-08T07:48:36Z)

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.