Fact-R1: Towards Explainable Video Misinformation Detection with Deep Reasoning

TL;DR

This paper introduces Fact-R1, a deep reasoning framework for video misinformation detection, supported by a large-scale benchmark, aiming to improve interpretability and reasoning capabilities over existing methods.

Contribution

It presents a novel deep reasoning framework, Fact-R1, combined with FakeVV, a large-scale dataset, advancing multimodal video misinformation detection with interpretability and reasoning.

Findings

Fact-R1 exhibits emergent reasoning behaviors similar to advanced RL systems.

FakeVV contains over 100,000 video-text pairs with detailed annotations.

The framework outperforms existing methods in interpretability and reasoning ability.

Abstract

The rapid spread of multimodal misinformation on social media has raised growing concerns, while research on video misinformation detection remains limited due to the lack of large-scale, diverse datasets. Existing methods often overfit to rigid templates and lack deep reasoning over deceptive content. To address these challenges, we introduce FakeVV, a large-scale benchmark comprising over 100,000 video-text pairs with fine-grained, interpretable annotations. In addition, we further propose Fact-R1, a novel framework that integrates deep reasoning with collaborative rule-based reinforcement learning. Fact-R1 is trained through a three-stage process: (1) misinformation long-Chain-of-Thought (CoT) instruction tuning, (2) preference alignment via Direct Preference Optimization (DPO), and (3) Group Relative Policy Optimization (GRPO) using a novel verifiable reward function. This enables Fact-R1 to exhibit emergent reasoning behaviors comparable to those observed in advanced text-based reinforcement learning systems, but in the more complex multimodal misinformation setting. Our work establishes a new paradigm for misinformation detection, bridging large-scale video understanding, reasoning-guided alignment, and interpretable verification.