Saved in:
Bibliographic Details
Main Authors: Sung, Junyoung, Lyu, Seungwoo, Kim, Minjun, An, Sumin, Nagrani, Arsha, Seo, Paul Hongsuck
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.01634
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911562476814336
author Sung, Junyoung
Lyu, Seungwoo
Kim, Minjun
An, Sumin
Nagrani, Arsha
Seo, Paul Hongsuck
author_facet Sung, Junyoung
Lyu, Seungwoo
Kim, Minjun
An, Sumin
Nagrani, Arsha
Seo, Paul Hongsuck
contents Real-world reasoning often requires combining information across modalities, connecting textual context with visual cues in a multi-hop process. Yet, most multimodal benchmarks fail to capture this ability: they typically rely on single images or set of images, where answers can be inferred from a single modality alone. This limitation is mirrored in the training data, where interleaved image-text content rarely enforces complementary, multi-hop reasoning. As a result, Vision-Language Models (VLMs) frequently hallucinate and produce reasoning traces poorly grounded in visual evidence. To address this gap, we introduce CRIT, a new dataset and benchmark built with a graph-based automatic pipeline for generating complex cross-modal reasoning tasks. CRIT consists of diverse domains ranging from natural images, videos, and text-rich sources, and includes a manually verified test set for reliable evaluation. Experiments on this benchmark reveal that even state-of-the-art models struggle on such reasoning tasks. Models trained on CRIT show significant gains in cross-modal multi-hop reasoning, including strong improvements on SPIQA and other standard multimodal benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2604_01634
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CRIT: Graph-Based Automatic Data Synthesis to Enhance Cross-Modal Multi-Hop Reasoning
Sung, Junyoung
Lyu, Seungwoo
Kim, Minjun
An, Sumin
Nagrani, Arsha
Seo, Paul Hongsuck
Machine Learning
Computation and Language
Real-world reasoning often requires combining information across modalities, connecting textual context with visual cues in a multi-hop process. Yet, most multimodal benchmarks fail to capture this ability: they typically rely on single images or set of images, where answers can be inferred from a single modality alone. This limitation is mirrored in the training data, where interleaved image-text content rarely enforces complementary, multi-hop reasoning. As a result, Vision-Language Models (VLMs) frequently hallucinate and produce reasoning traces poorly grounded in visual evidence. To address this gap, we introduce CRIT, a new dataset and benchmark built with a graph-based automatic pipeline for generating complex cross-modal reasoning tasks. CRIT consists of diverse domains ranging from natural images, videos, and text-rich sources, and includes a manually verified test set for reliable evaluation. Experiments on this benchmark reveal that even state-of-the-art models struggle on such reasoning tasks. Models trained on CRIT show significant gains in cross-modal multi-hop reasoning, including strong improvements on SPIQA and other standard multimodal benchmarks.
title CRIT: Graph-Based Automatic Data Synthesis to Enhance Cross-Modal Multi-Hop Reasoning
topic Machine Learning
Computation and Language
url https://arxiv.org/abs/2604.01634