Independent eval of Openai/privacy-filter vs GLiNER on 600 PII samples. The model is much better than naive benchmarks make it look
·
0 reactions
·
0 comments
·
3 views
OpenAI dropped Privacy Filter last month under Apache 2.0 and I wanted to see how it actually stacks up against the other serious open weight option for PII detection, GLiNER large-v2.1. Ran a full head to head on 600 labeled samples from ai4privacy (400 English, 200 across French, German, Spanish, Italian, Dutch). The headline finding is that openai/privacy-filter is genuinely strong, but you'd never know it from a quick benchmark. Here's why: Openai/privacy-filter is a token classifier with a
Original article
OpenAI
Anonymous · no account needed
