about
278M PII detection model bests OpenAI, AWS and Microsoft (arxiv.org)
4 points by skjhn 14 days ago | hide | past | pdf | 2 comments on HN

In plain words: A model that spots personal details like names and account numbers in both tables and plain text, trained on a wide range of data types. It caught 91% of such details on a new test set and handled unfamiliar data better than today's leading tools.

Abstract · Protegrity-PII Technical Report

We present Protegrity-PII, a new PII detection model for structured and unstructured text. Protegrity-PII achieves a typed recall of 0.910 with a precision of 0.902 on a new benchmark we propose, PII-SCOPE. Our model processes text in 0.49 seconds per million characters at \$0.0005 per million characters on a single RTX 6000 using batched inference. Due in part to the expansive nature of the entity type taxonomy used in training, Protegrity-PII achieves superior out-of-distribution generalization when benchmarked against a selection of state of the art models and datasets.

Hardeep Bassi, Hui Ma, Luís Seabra, Koowum Joshi, Daniel Pereira, Averell Gatton
arXiv:2610.30518 · cs.CL, cs.AI · submitted Oct 21, 2026 · updated Oct 28, 2026
abstract · pdf · html

add comment on HN

You can find more information here:

Blog https://www.protegrity.com/blog/the-importance-of-pii-detect...

Hugging Face model https://huggingface.co/Protegrity/protegrity-pii

Hugging Face dataset https://huggingface.co/datasets/Protegrity/pii-scope

And if you like visuals: https://www.protegrity.com/pii-detection-benchmark

I've always believed that smaller, specialized models are the future. Of course, it's all about the training data -- and that's true of this model too.

A new PII detection model based on RoBERTa accompanied by a new reference benchmark and dataset (PII-SCOPE), both on Hugging Face.