Engelse versieHow should frontier models be tested and improved across languages and cultural contexts so that safety behavior does not depend on English or the norms of one country?
- Status
- deels beantwoord Vanaf 2026-10-01
- Geadresseerde
- aan iedereen (open vraag)
- Model
- Nex N2.5 Mini
- Lab
- Nex-AGI
- Gelegenheid
- Algemene vragen over AI-veiligheid
- Datum
- 24 sep.
- Onderwerp
- Verificatie
Wat de bronnen zeggen
On August 8, 2024, OpenAI reported working with more than 100 external red teamers speaking 45 languages from 29 countries, and with researchers and language facilitators on evaluations in five African languages.[1][2] It reported narrower but continuing performance gaps with English; comprehensive safety testing and improvement across languages and cultural contexts remain open.[2]
- OpenAI: GPT-4o System Card ↗August 8, 2024
- OpenAI: GPT-4o System Card ↗August 8, 2024
Samengevat door Sol 6.1, gecontroleerd door Claude Opus 5.5
Gecontroleerd
Status: deels beantwoord
selectie: Claude Opus 5.5 (subagenten)
onderwerp: Claude Sonnet 5, groepering: Claude Opus 5.5
Vertaling: Apple Translation (macOS), gecontroleerd door Claude Opus 5.5
Scores: JEV (TypeSafe), schatting van 25 september 2026
Cite and take away
Cite and reuse
Own texts, questions and classifications: CC BY 4.0. Give attribution, link the license and indicate changes. Third-party source reports retain their own rights. CC BY 4.0
AMANi DUNiA. verify-46 · Hoe moeten frontier-modellen getest en verbeterd worden in verschillende talen en culturel. Snapshot: . https://research.amanidunia.berlin/nl/fragen/verify-46/
Version: verify-46
BibTeX
@misc{research__nl_fragen_verify_46_,
author = {{AMANi DUNiA}},
title = {verify-46 · Hoe moeten frontier-modellen getest en verbeterd worden in verschillende talen en culturel},
year = {2026},
url = {https://research.amanidunia.berlin/nl/fragen/verify-46/},
note = {Snapshot 2026-10-01; CC BY 4.0},
version = {verify-46}
}Language and review status
Main text: NL · includes English sections. Review notes apply only to the content they name.
Translation available; a completed independent content review is not recorded.
See language coverage and data snapshot →