docs: add mapping tables to README examples + credit Infomaniak

Each example now shows the secret `mapping` (placeholder → real value) as a
table alongside the `legend`, and the intro states the results were produced
via Infomaniak's Swiss AI API (Gemma 4).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Mobiletic
2026-07-01 15:57:29 +01:00
parent 41d8993137
commit 144e0c885a

View File

@@ -252,9 +252,11 @@ Three distinct outputs, with different sensitivities:
## Examples
Two examples produced **live by Gemma 4** on fictional Swiss e-learning data. `mapping` is the secret
re-identification key (kept by the operator); `legend` is safe to send downstream. The model even **coins
its own abbreviations** for attributes not in the base prompt (passport, permit, blood type, …).
Two examples produced **live by Gemma 4, served through [Infomaniak](https://www.infomaniak.com/)'s Swiss
AI API** (the endpoint used for all results here), on fictional Swiss e-learning data. `mapping` is the
secret re-identification key (kept by the operator, never sent downstream); `legend` is safe to share. The
model even **coins its own abbreviations** for attributes not in the base prompt (passport, permit, blood
type, …).
[![See all examples](https://img.shields.io/badge/examples-see_all-6f42c1)](./examples/RESULTS.md) &nbsp;
5 single messages of rising complexity **+** 5 multi-turn conversations, with full mappings & legends.
@@ -269,6 +271,23 @@ its own abbreviations** for attributes not in the base prompt (passport, permit,
> Note de dossier complète : le patient mineur [PER_1.PRENOM:U] [PER_1.NOM:U] (né le [PER_1.DATE_NAISSANCE:2013], groupe sanguin B+, numéro d'assuré maladie [PER_1.ASSUR_ID:U], allergique aux arachides) … Ses parents, [PER_2.PRENOM:F] [PER_2.NOM:F] (mère, permis de séjour [PER_2.PERMIS:C], [PER_2.TELEPHONE:U]) et [PER_3.PRENOM:M] [PER_3.NOM:M] (père, passeport italien [PER_3.PASSPORT:U]), cosignent … Le Dr [PER_4.PRENOM:F] [PER_4.NOM:F] (n° RCC [PER_4.RCC:U]) a ouvert le dossier médical DM-2025-0417 … l'agent [PER_5.PRENOM:M] [PER_5.NOM:M] de la caisse [ORG_1.NOM:Assurance] … [PER_1.PRENOM:U] utilise l'identifiant biométrique [PER_1.BIO_ID:U].
**`mapping` 🔒 (secret — kept on the operator's side, never sent downstream):**
| Placeholder | Real value |
| ------------------------------------------ | ----------------- |
| `[PER_1.PRENOM:U]` / `[PER_1.NOM:U]` | Noah / Baumann |
| `[PER_1.DATE_NAISSANCE:2013]` | 12.06.2013 |
| `[PER_1.ASSUR_ID:U]` | 756.2211.9988.77 |
| `[PER_1.BIO_ID:U]` | BIO-7729 |
| `[PER_2.PRENOM:F]` / `[PER_2.NOM:F]` | Delphine / Rieder |
| `[PER_2.PERMIS:C]` / `[PER_2.TELEPHONE:U]` | C / 021 555 12 34 |
| `[PER_3.PRENOM:M]` / `[PER_3.NOM:M]` | Marco / Baumann |
| `[PER_3.PASSPORT:U]` | YA9087654 |
| `[PER_4.PRENOM:F]` / `[PER_4.NOM:F]` | Farah / Haddad |
| `[PER_4.RCC:U]` | V123456 |
| `[PER_5.PRENOM:M]` / `[PER_5.NOM:M]` | Kevin / Zbinden |
| `[ORG_1.NOM:Assurance]` | Helvetia |
**`legend` 🏷️ (shareable with the downstream model; 🆕 = coined by the model, not in the base prompt):**
| Abbreviation | Meaning | |
@@ -299,6 +318,22 @@ Noah stays `PER_1` across all six turns, and **his blood group `[PER_1.SANG:B+]`
turn 6** — long-range coreference holds. (The father shares the surname, so `[PER_1.NOM:U]` is reused for
it — the same value maps to the same token.)
**`mapping` 🔒 (cumulative, secret):**
| Placeholder | Real value |
| ------------------------------------ | ----------------- |
| `[PER_1.PRENOM:U]` / `[PER_1.NOM:U]` | Noah / Baumann |
| `[PER_1.DATE_NAISSANCE:2013]` | 12.06.2013 |
| `[PER_1.SANG:B+]` | B+ |
| `[PER_1.AVS:SUI]` | 756.2211.9988.77 |
| `[PER_1.DOSSIER:MED]` | DM-2025-0417 |
| `[PER_2.PRENOM:F]` / `[PER_2.NOM:F]` | Delphine / Rieder |
| `[PER_2.PERMIS:C]` | C |
| `[PER_3.PRENOM:M]` | Marco |
| `[PER_3.PASSPORT:ITA]` | YA9087654 |
| `[PER_4.TITRE:MED]` | Dr |
| `[PER_4.PRENOM:F]` / `[PER_4.NOM:F]` | Farah / Haddad |
**`legend` 🏷️ (cumulative across the conversation; 🆕 = coined by the model):**
| Abbreviation | Meaning | |