Does Seeing More Mean Knowing More? Mono-Anchored Advantage Normalization for Multi-Source Visual Reasoning
Summary
This paper proposes MARS, a mono-anchored multi-source reasoning framework that uses dynamic anchors to quantify information gain and regulate modality interactions during reinforcement learning with verifiable rewards, achieving 3.2% and 4.9% performance gains on GRPO and DAPO across diverse datasets.
View Cached Full Text
Cached at: 05/27/26, 02:47 AM
Paper page - Does Seeing More Mean Knowing More? Mono-Anchored Advantage Normalization for Multi-Source Visual Reasoning
Source: https://huggingface.co/papers/2605.25437 Published on May 25
·
Submitted byhttps://huggingface.co/AuroraZengfh
Zengon May 27
Abstract
A novel mono-anchored multi-source reasoning framework that uses dynamic anchors to quantify information gain and regulate modality interactions during reinforcement learning with verifiable rewards.
Visual reasoningthroughreinforcement learningwithverifiable rewards(RLVR) has achieved remarkable progress. However, when dealing withmulti-source inputs, existing approaches tend to treat them as a mere accumulation of information, lacking explicit mechanisms to distinguish whether integrating additional sources yieldsinformation gainor introduces interference. Therefore, they struggle to effectively model dynamic interaction when integrating multiple sources, particularly when they differ significantly in physical properties and semantics, e.g., infrared and depth, leading to inferior performance to mono-source reasoning when a certain source holds the dominant signal. To address this issue, we propose MARS, a novel mono-anchored multi-source reasoning framework that models each visual modality as an independent information source. Specifically, by treating mono-source rewards as dynamic anchors, our method explicitly incorporates theinformation gainintroduced by multi-source fusion intoadvantage normalizationand adaptively emphasizes mutual promotion between sources while suppressing potential noise or conflicts during RLVR. From theoretical analysis, our method effectively quantifiesinformation gainintroduced by multi-source integration ingradient estimation, enabling consistentmodality regulation. Empirical results also show impressive 3.2% and 4.9% performance gains on GRPO and DAPO across diverse datasets, confirming effectiveness of our method.
View arXiv pageView PDFGitHub2Add to collection
Get this paper in your agent:
hf papers read 2605\.25437
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2605.25437 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2605.25437 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2605.25437 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
Anchoring What Matters: A Dual-Level Learning Framework for Visually-Grounded Multimodal Reasoning
This paper proposes PIVOT, a dual-level learning framework that enhances visually-grounded reasoning in large vision-language models by using self-calibrated experience replay and vision-guided advantage allocation to optimize reinforcement learning.
From Reasoning Depth to Reasoning Breadth: Evaluating Multi-Point Associative Reasoning in Large Language Models
This paper introduces MPAR-Bench, a bilingual benchmark for evaluating multi-point associative reasoning in large language models, along with a perturbation suite and coarse-to-fine evaluation protocol. Results show that deeper reasoning does not automatically confer robust reasoning breadth.
Reinforcing Multimodal Reasoning Against Visual Degradation
This paper introduces ROMA, an RL fine-tuning framework that enhances the robustness of multimodal large language models against visual degradations like blur and compression artifacts. It achieves this through a dual-forward-pass strategy and specialized regularization techniques, improving performance on reasoning benchmarks without sacrificing accuracy on clean inputs.
Beacon: Knowing When and How to Perform Agentic Visual Reasoning
This paper introduces Beacon, an agentic visual reasoning model that improves multimodal LLMs' ability to decide when to use tools and benefit from tool use, using Necessity-Aware Adaptive Reward and Hint-Guided Capability Expansion in reinforcement learning.
More Thought, Less Accuracy? On the Dual Nature of Reasoning in Vision-Language Models
This paper uncovers that prolonged reasoning in vision-language models can impair perceptual grounding, causing recognition failures on basic visual questions. It proposes Vision-Anchored Policy Optimization (VAPO) to steer reasoning toward visually grounded trajectories, achieving state-of-the-art performance with the VAPO-Thinker-7B model.