Seeing with You: Perception-Reasoning Coevolution for Multimodal Reasoning

TL;DR

PRCO introduces a dual-role reinforcement learning framework for multimodal reasoning, improving visual evidence extraction and reasoning accuracy by coordinating an Observer and Solver with role-specific rewards.

Contribution

It proposes PRCO, a novel perception-reasoning coevolution approach with shared policy and role-specific rewards, enhancing multimodal reasoning performance.

Findings

PRCO improves accuracy by over 7 points across benchmarks.

PRCO outperforms prior open-source RL-tuned baselines.

PRCO demonstrates consistent gains across model scales.

Abstract

Reinforcement learning with verifiable rewards (RLVR) has substantially enhanced the reasoning capabilities of multimodal large language models (MLLMs). However, existing RLVR approaches typically rely on outcome-driven optimization that updates both perception and reasoning using a shared reward based solely on the final answer. This shared reward blurs credit assignment, frequently improving reasoning patterns while failing to reliably enhance the accuracy of upstream visual evidence extraction. To address this perception bottleneck, we introduce PRCO (Perception-Reasoning Coevolution), a dual-role RLVR framework with a shared policy. PRCO consists of two cooperative roles: an Observer that generates an evidence caption tailored to the question and a Solver that predicts the final answer based on this caption. Crucially, PRCO employs role-specific reward signals: the Solver is optimized using verifiable outcome rewards on the final answer, while the Observer receives a utility reward derived from the Solver's downstream success. Extensive experiments across eight challenging multimodal reasoning benchmarks demonstrate that PRCO yields consistent improvements across model scales by over 7 points on average accuracy compared to the base model, outperforming prior open-source RL-tuned baselines.