Emergent Introspective Awareness in Large Language Models

(arxiv.org)

12 points | by doener 2 hours ago ago

3 comments

  • Retro_Dev an hour ago

    > Overall, our results indicate that current language models possess some functional introspective awareness of their own internal states. We stress that in today’s models, this capacity is highly unreliable and context-dependent; however, it may continue to develop with further improvements to model capabilities.

    • Escapade5160 8 minutes ago

      If it's context dependent is it really introspection?

  • whythismatters an hour ago

    >Submitted on 5 Jan 2026