Gemini glitch leaks its backend response instructions to the user in plain text
SushiManPlatinum · reddit · 2026-11-13
A Reddit user caught Gemini glitching and displaying the backend instructions meant for the model. The screenshot shows internal guidance: match the user's blunt "fuck that" energy without escalating or lecturing, keep it short and plain, and include a low-pressure <FollowUp> as required by the system prompt. The leak offers a rare look at how Google engineers the assistant's tone and follow-up behavior.
More from Models
- AI Daily: Huang resigns, Anthropic IPO at up to $1.4T, Gemini 4 Carbon opens — xiaohu · 2026-11-13
- Dylan Patel: Moonshot and DeepSeek's open-source code only targets Blackwell — anshulkundaje · 2026-11-13
- Tesla FSD handles a dark, snowy Finnish road and yields to a bus smoothly — yunta_tsai · 2026-11-13
- Claude now refuses to work on open-source projects that ban AI contributions — esrtweet · 2026-11-13
- Thinking Machines' Inkling 2 Small (594B MoE, 1M context) lands on Ollama — ollama · 2026-11-13
- No forensic evidence DeepSeek's V4.1-Flash used Blackwell, buried in report appendix — suchenzang · 2026-11-13