Patronus AI’s Judge-Image wants to keep AI honest — and Etsy is already using it
1 min read
Summary
Patronus AI has launched an MLMM-as-a-judge, called Judge-Image, which is an AI tool evaluating systems that interpret images and produce text, with the aim of detecting and mitigating hallucinations and reliability issues.
Etsy has already implemented this technology to check the accuracy of captions for product images on its site.
Patronus chose Google’s Gemini model over OpenAI’s GPT-4V as the underlining model for its MLLM judge.
The company offers a free option which allows users to try the platform up to a certain limit, after which customers pay for evaluator usage or can engage enterprise options.
The company plans to expand into audio evaluation in the future.