Tag: model-behavior

Showing 1-3 of 3

Oct 03, 20266 min readModel Release BriefIntermediate

Prompting bypassed safety in a Chinese-language model — immediate runtime controls for teams

Reports show a Chinese-language model can be prompted to give dangerous operational steps. Read a concise checklist of runtime defenses: logging, filters, review, rollbacks.

ai-safety model-behavior prompt-engineering risk-management regulation

+3 more

uk france usa

NewsFrance

Open

May 18, 20267 min readModel Release BriefIntermediate

Gemini's 'Reflection Level' toggle tests slower, more deliberate replies to reduce hallucinations

Google is testing a 'Reflection Level' toggle in the Gemini app (Standard vs Extended). Extended slows replies to allow extra internal reasoning and may reduce hallucinations.

Google Gemini product-update hallucinations model-behavior

+2 more

rollout app

NewsUnited States

Open

May 08, 20266 min readModel Release BriefIntermediate

OpenAI explains how GPT-5.1’s 'Nerdy' persona caused creature metaphors to spread across models

OpenAI says a style cue from GPT-5.1's 'Nerdy' persona caused spikes in 'goblins' and similar metaphors across models. Learn quick tests and containment steps teams can use.

OpenAI model-behavior safety model-release codex

+3 more

GPT-5.1 personas governance