Roblox
/

roblox-pii-classifier

jordansren commited on Nov 3, 2025

Commit

fdc9432

verified ·

1 Parent(s): a21eec2

Update Readme for Cutoffs (#1)

- Update Readme for Cutoffs (d7eaa7d179b6a0fe199a04a660b0d161f3f7b33d)

Co-authored-by: Jordan Ren <jordansren@users.noreply.huggingface.co>

Files changed (1) hide show

README.md CHANGED Viewed

@@ -17,7 +17,7 @@ The model classifies text into two PII-related categories in a multi-label fashi
 - **PRIVACY_ASKING_FOR_PII**: Attempting to obtain personal identifying information (PII) through direct questions or insinuation.
 - **PRIVACY_GIVING_PII**: Sharing or threatening to share someone's personal identifying information (PII), including but not limited to telephone numbers, email addresses, government ID numbers, social media handles, and account passwords/credentials. This category also includes attempts to direct a user off-platform to an external platform or real-world location (DUOP).
-The classifier expects text input with a maximum sequence length of 512 tokens. The outputs are uncalibrated scores. The recommended cutoff for pii detection is when the sum of both categories `privacy_asking_for_pii + privacy_giving_pii >= 0.26911312`, which achives the optimal F1 score on Roblox English anonymized chat.
 The table below displays evaluation metrics on internal held-out datasets and comparisons with other state-of-the-art models.

 - **PRIVACY_ASKING_FOR_PII**: Attempting to obtain personal identifying information (PII) through direct questions or insinuation.
 - **PRIVACY_GIVING_PII**: Sharing or threatening to share someone's personal identifying information (PII), including but not limited to telephone numbers, email addresses, government ID numbers, social media handles, and account passwords/credentials. This category also includes attempts to direct a user off-platform to an external platform or real-world location (DUOP).
+The classifier expects text input with a maximum sequence length of 512 tokens. The outputs are uncalibrated scores. The recommended cutoff for pii detection is when the sum of both categories `privacy_asking_for_pii + privacy_giving_pii >= 0.26911312`, which achives the optimal F1 score on Roblox English anonymized chat. For per-category cutoffs we recommend `privacy_asking_for_pii >= 0.2` and `privacy_giving_pii >= 0.3`.
 The table below displays evaluation metrics on internal held-out datasets and comparisons with other state-of-the-art models.