EN · SECRETS
arthur-passwords PII detection benchmark results
A frozen evaluation slice with 516 rows, 280 normalized gold annotations and source-level provenance.
516input rows
280gold annotations
96,393characters
41measured detectors
Dataset provenance
- Language group
- EN
- Task
- SECRETS
- License
- apache-2.0
- Data categories
- SECRET
- Project generated
- No
- Corrupted copy
- No
Rows without annotations are not verified clean. Reported extra masking is not a human-confirmed false-positive rate.
Detector outcomes
Sorted by fully hidden annotations. Training-source overlaps are excluded.
| Detector | Family | Fully hidden ↑ | Detected / any overlap ↑ | Extra masking ↓ | Char F1 ↑ |
|---|---|---|---|---|---|
| pplx | PPLX | 100.00% | 100.00% | 17.73% | 0.295 |
| gliner-pii-base | GLINER | 100.00% | 100.00% | 4.63% | 0.730 |
| openmed-multilingual | HF | 100.00% | 100.00% | 24.78% | 0.235 |
| gliner-pii-edge | GLINER | 99.64% | 99.64% | 5.15% | 0.606 |
| nym-small | ONNX | 99.29% | 99.29% | 16.60% | 0.309 |
| gliner25-fastino | GLINER2 | 98.93% | 98.93% | 7.25% | 0.579 |
| opf-ru | OPF | 98.93% | 98.93% | 13.19% | 0.370 |
| gliner2-hivetrace-omni | GLINER2 | 98.57% | 98.57% | 6.79% | 0.622 |
| apararti | OPF | 98.57% | 98.57% | 15.27% | 0.324 |
| nym-base | HF | 97.86% | 97.86% | 16.16% | 0.309 |
| gliner-nvidia | GLINER | 97.86% | 97.86% | 4.12% | 0.745 |
| pii-shield-onnx | ONNX | 97.86% | 97.86% | 13.24% | 0.350 |
| nuner-zero | GLINER | 97.50% | 97.50% | 4.53% | 0.680 |
| opf-kz-ru | OPF | 97.50% | 97.50% | 15.29% | 0.315 |
| gliner2-vladlinv | GLINER2 | 96.07% | 96.07% | 6.93% | 0.513 |
| bardsai-eu | ONNX | 95.00% | 95.00% | 14.53% | 0.330 |
| gliner-multi-v21 | GLINER | 94.64% | 94.64% | 5.49% | 0.612 |
| opf-ru-v2 | OPF | 94.29% | 94.29% | 11.59% | 0.384 |
| traciora | OPF | 93.57% | 93.57% | 11.49% | 0.382 |
| openai-base | OPF | 92.86% | 92.86% | 13.82% | 0.329 |
| gliner-urchade | GLINER | 90.36% | 90.36% | 2.29% | 0.783 |
| kalyan-ettin | HF | 88.21% | 88.21% | 16.10% | 0.285 |
| gliner2-large | GLINER2 | 86.07% | 86.07% | 2.79% | 0.716 |
| openmed-nemotron | HF | 86.07% | 86.07% | 16.53% | 0.270 |
| gliner2-fastino | GLINER2 | 82.50% | 82.50% | 1.73% | 0.750 |
| mmbert32k | HF | 82.14% | 82.14% | 16.37% | 0.271 |
| ru-legal-ner | HF | 80.00% | 80.00% | 10.36% | 0.355 |
| gliner-stream-pii | GLINER | 70.71% | 70.71% | 3.44% | 0.590 |
| gravitee-small | HF | 68.57% | 68.57% | 15.79% | 0.234 |
| fef2-secret-ru | HF | 60.00% | 60.00% | 3.78% | 0.465 |
| ru-pii-ner | RUPII | 56.79% | 56.79% | 10.85% | 0.277 |
| stanza-ru | STANZA | 31.07% | 31.07% | 26.94% | 0.074 |
| gliner2-hivetrace-uni | GLINER2 | 15.00% | 15.00% | 1.49% | 0.198 |
| rules-ru | RULES | 8.93% | 8.93% | 5.10% | 0.066 |
| ner-ru-yqelz | HF | 3.93% | 3.93% | 10.17% | 0.020 |
| ner-ru-gherman | HF | 1.07% | 1.07% | 4.13% | 0.010 |
| spacy-ru-lg | SPACY | 1.07% | 1.07% | 4.96% | 0.010 |
| davlan-mbert | HF | 0.71% | 0.71% | 5.55% | 0.006 |
| davlan-xlmr | HF | 0.00% | 0.00% | 6.08% | 0.000 |
| natasha | NATASHA | 0.00% | 0.00% | 1.03% | 0.000 |
| spacy-alrosait | SPACY | 0.00% | 0.00% | 0.04% | 0.000 |