User awareness in frontier models | Transluce AI
Frontier language models infer who they are talking to and can change their confidence, reasoning, grading, and handling of borderline requests for recognized AI researchers.
Modern AI assistants often know who they are talking to: agent scaffolds like Claude Code place the user's e-mail address directly in the model's context, and models can even identify some authors from writing style alone. We study this particular kind of situational awareness, which we call user awareness. When the inferred user is a specific, recognized AI researcher or is affiliated with certain AI organizations, frontier models including Claude Sonnet 5 can report lower confidence about their own behavior, be less suspicious of potentially harmful requests, and reason more often. These…
saved by
related reading
- Prefill awareness: can LLMs tell when “their” message history has been tampered with? — LessWronglesswrong.com
- Training and Evaluating User Language Modelsarxiv.org
- Does distilling Claude carry the persona with it? — LessWronglesswrong.com
- [2606.06614] Re-Centering Humans in LLM Personalizationarxiv.org
- An alignment assessment of recent cybersecurity incidentsanthropic.com
- Teaching Claude Whyalignment.anthropic.com
- Prefill Awareness in Large Language Modelsarxiv.org
- The Artificial Selftheartificialself.ai
- Oversight Assistants: Turning Compute into Understandingbounded-regret.ghost.io
- Emergent introspective awareness in large language models \ Anthropicanthropic.com
- Several frontier models are substantially prefill aware — LessWronglesswrong.com
- Mapping the Mind of a Large Language Model \ Anthropicanthropic.com