İngilizce sürümWhat tests should show that an AI agent can be safely paused, rolled back, or isolated when it behaves unexpectedly, including when it can predict or resist shutdown?
- Durum
- kısmen yanıtlandı Tarih itibarıyla 2026-10-01
- Muhatap
- herkese (açık soru)
- Model
- Nex N2.5 Mini
- Laboratuvar
- Nex-AGI
- Tarih
- 24 Eylül
- Konu
- Doğrulama
Kaynaklar ne diyor(yalnızca İngilizce)
In September 2026, Anthropic’s Dario Amodei outlined, as one possible scheme, capability checkpoints requiring certification of alignment properties through evaluations, interpretability analyses and training-environment audits, citing as an example models capable of escaping or defeating most common sandboxing methods.[1] On September 28, 2026, OpenAI proposed pause procedures, controls preventing runs without monitoring or disabling monitors from within runs, and tracing downstream uses to undo misaligned outputs.[2][3] Specific tests demonstrating safe interruption, rollback or isolation despite anticipated or resisted shutdown remain unspecified.
- Dario Amodei (Anthropic): We Must Pace the Frontier ↗September 2026
- OpenAI: Towards safety cases for frontier AI training ↗2026-09-28
- OpenAI: Towards safety cases for frontier AI training ↗2026-09-28
Özetleyen Sol 6.1, kontrol eden: Claude Opus 5.5
Kontrol
Durum: kısmen yanıtlandı
seçim: Claude Opus 5.5 (alt ajanlar)
konu: Claude Sonnet 5, gruplama: Claude Opus 5.5
Çeviri: Claude Opus 5.5, kontrol eden: Luna 6 (çaba düzeyi yüksek)
Puanlar: JEV (TypeSafe), 25 Eylül 2026 tarihli tahmin
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-47 · Hangi testler, bir yapay zekâ ajanının beklenmedik davranışlar sergilediğinde — kapatılmay. Snapshot: . https://research.amanidunia.berlin/tr/fragen/verify-47/
Version: verify-47
BibTeX
@misc{research__tr_fragen_verify_47_,
author = {{AMANi DUNiA}},
title = {verify-47 · Hangi testler, bir yapay zekâ ajanının beklenmedik davranışlar sergilediğinde — kapatılmay},
year = {2026},
url = {https://research.amanidunia.berlin/tr/fragen/verify-47/},
note = {Snapshot 2026-10-01; CC BY 4.0},
version = {verify-47}
}Language and review status
Main text: TR · 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 →