# Multi-Stage Verification-Centric Framework for Mitigating Hallucination in Multi-Modal RAG

arXiv: https://arxiv.org/abs/2507.20136 · HTML: https://mlsanity.com/paper/2507.20136/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 3 of its 11 references that are in the map.
- First posted on arXiv 27 Jul 2025.

## Builds on (most cited first, 3 of 3)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation](https://mlsanity.com/paper/2402.03216/) (2024)

## Built on by (most cited first, 5 of 5)

- [DynHD: Hallucination Detection for Diffusion Large Language Models via Denoising Dynamics Deviation Learning](https://mlsanity.com/paper/2603.16459/) (2026)
- [DeMTS: Denoising Trajectories as Multivariate Time Series for Hallucination Detection in Diffusion Language Models](https://mlsanity.com/paper/2608.14632/) (2026)
- [SIM: Subspace Interaction-based Method for Token-Level Text Anomaly Detection](https://mlsanity.com/paper/2609.08200/) (2026)
- [HOPE: Heterophily-Aware Open-Set Node Classification with Pseudo-Extrapolation](https://mlsanity.com/paper/2609.08685/) (2026)
- [DiFA: Dual Evidence Fusion and Aggregation for Token-Level Text Anomaly Detection](https://mlsanity.com/paper/2609.22136/) (2026)

## Abstract (arXiv)

This paper presents the technical solution developed by team CRUISE for the KDD Cup 2025 Meta Comprehensive RAG Benchmark for Multi-modal, Multi-turn (CRAG-MM) challenge. The challenge aims to address a critical limitation of modern Vision Language Models (VLMs): their propensity to hallucinate, especially when faced with egocentric imagery, long-tail entities, and complex, multi-hop questions. This issue is particularly problematic in real-world applications where users pose fact-seeking queries that demand high factual accuracy across diverse modalities. To tackle this, we propose a robust, multi-stage framework that prioritizes factual accuracy and truthfulness over completeness. Our solution integrates a lightweight query router for efficiency, a query-aware retrieval and summarization pipeline, a dual-pathways generation and a post-hoc verification. This conservative strategy is designed to minimize hallucinations, which incur a severe penalty in the competition's scoring metric. Our approach achieved 3rd place in Task 1, demonstrating the effectiveness of prioritizing answer reliability in complex multi-modal RAG systems. Our implementation is available at https://github.com/Breezelled/KDD-Cup-2025-Meta-CRAG-MM .