生成式通用驗證器作為多模態元推理器

Generative Universal Verifier as Multimodal Meta-Reasoner

October 15, 2025

作者: Xinchen Zhang, Xiaoying Zhang, Youbin Wu, Yanbin Cao, Renrui Zhang, Ruihang Chu, Ling Yang, Yujiu Yang

cs.AI

摘要

我們提出了生成式通用驗證器這一新穎概念及插件，專為下一代視覺語言模型和統一多模態模型中的多模態推理而設計，在推理和生成過程中提供了對視覺結果進行反思與精煉的基礎能力。本工作主要貢獻有三：(1) 我們構建了ViVerBench，這是一個涵蓋16類關鍵任務的綜合基準，用於評估多模態推理中的視覺結果。結果顯示，現有的視覺語言模型在這些任務上普遍表現不佳，凸顯了與人類在可靠視覺驗證能力上的顯著差距。(2) 我們設計了兩條自動化流水線來構建大規模視覺驗證數據，並訓練了OmniVerifier-7B，這是首個為通用視覺驗證訓練的全能生成式驗證器，在ViVerBench上取得了顯著提升(+8.3)。通過訓練，我們識別了視覺驗證中的三項基本能力，並展示了它們如何協同泛化與互動。(3) 我們提出了OmniVerifier-TTS，這是一種序列化測試時擴展範式，利用通用驗證器在統一模型內橋接圖像生成與編輯，通過迭代的細粒度優化提升生成能力的上限。除生成外，我們還將通用驗證器擴展至更廣泛的世界建模交錯推理場景。實證表明，OmniVerifier-TTS在T2I-ReasonBench(+3.7)和GenEval++(+4.3)上取得了改進，超越了現有的並行測試時擴展方法，如Best-of-N。通過賦予多模態推理可靠的視覺驗證能力，OmniVerifier不僅提升了生成過程中的可靠反思，還實現了可擴展的測試時精煉，標誌著向更可信、更可控的下一代推理系統邁進了一步。

English

We introduce Generative Universal Verifier, a novel concept and plugin designed for next-generation multimodal reasoning in vision-language models and unified multimodal models, providing the fundamental capability of reflection and refinement on visual outcomes during the reasoning and generation process. This work makes three main contributions: (1) We build ViVerBench, a comprehensive benchmark spanning 16 categories of critical tasks for evaluating visual outcomes in multimodal reasoning. Results show that existing VLMs consistently underperform across these tasks, underscoring a substantial gap from human-level capability in reliable visual verification. (2) We design two automated pipelines to construct large-scale visual verification data and train OmniVerifier-7B, the first omni-capable generative verifier trained for universal visual verification and achieves notable gains on ViVerBench(+8.3). Through training, we identify three atomic capabilities in visual verification and demonstrate how they generalize and interact synergistically. (3) We propose OmniVerifier-TTS, a sequential test-time scaling paradigm that leverages the universal verifier to bridge image generation and editing within unified models, enhancing the upper bound of generative ability through iterative fine-grained optimization. Beyond generation, we extend universal verifier to broader world-modeling interleaved reasoning scenarios. Empirically, OmniVerifier-TTS achieves improvements on T2I-ReasonBench(+3.7), and GenEval++(+4.3), outperforming existing parallel test-time scaling methods, such as Best-of-N. By endowing multimodal reasoning with reliable visual verification, OmniVerifier advances both reliable reflection during generation and scalable test-time refinement, marking a step toward more trustworthy and controllable next-generation reasoning systems.