Additional Details on On-Device Medical Data Anonymization
MaziyarPanahi · x · 2026-07-13
The author further explains that this local anonymization model identifies and removes various identifiers on-device, including names, medical record numbers, social security numbers, dates of birth, phone numbers, emails, addresses, and policy numbers.
Its design goal is not semantic rewriting, but rather de-identification while preserving the content's meaning. This same ONNX solution can also run on Android via ONNX Runtime, and the model is released under the Apache-2.0 license.
Related event: HF Expands Edge ONNX as Medical Redaction Goes Local(3 posts)→
More from Infra
- Tabul AI launches Metal TreeSHAP to speed up Shapley values on Apple silicon — Scobleizer · 2026-07-22
- DeepSeek-V4-Flash tops out at 770 tok/s on one B300 in a vLLM batch test — Moreh · 2026-07-22
- NVIDIA starts shipping 102.4 Tbps Spectrum-6 switches for Vera Rubin AI factories — nvidia · 2026-07-22
- Apple publishes SOC 3 audit reports for Private Cloud Compute — throwfaraway4 · 2026-07-22
- Reddit GPU renters say existing platforms only give you two of three: code, recovery, fair billing — legendpizzasenpai · 2026-07-22
- The Sandboxing Manifesto: Secure Execution Environments for Agents — spirosoik · 2026-07-22