Emergent introspective awareness in large language models \ Anthropic

Research from Anthropic on the ability of large language models to introspect

Interpretability Signs of introspection in large language models Oct 29, 2025 Read the paper Have you ever asked an AI model what’s on its mind? Or to explain how it came up with its responses? Models will sometimes answer questions like these, but it’s hard to know what to make of their answers. Can AI systems really introspect—that is, can they consider their own thoughts? Or do they just make up plausible-sounding answers when they’re asked to do so? Understanding whether AI systems can truly introspect has important implications for their transparency and reliability. If models can accurat

saved by

• Lydia Nottingham
• mateus de sousa
• Will Anderson
• Harrison Qian
• jenny peng

related reading

• Tracing the thoughts of a large language model \ Anthropicanthropic.com
• Emergent Introspective Awareness in Large Language Modelstransformer-circuits.pub
• Emergent Introspection in AI is Content-Agnosticarxiv.org
• Can LLMs Introspect? A Reality Checkarxiv.org
• Emergent Introspective Awareness in Large Language Modelstransformer-circuits.pub
• [2508.14802] Privileged Self-Access Matters for Introspection in AIarxiv.org
• Discovering Language Model Behaviors with Model-Written Evaluations — LessWronglesswrong.com
• Small Models Can Introspect, Toovgel.me
• [2601.01828] Emergent Introspective Awareness in Large Language Modelsarxiv.org
• A global workspace in language models \ Anthropicanthropic.com
• On the Biology of a Large Language Modeltransformer-circuits.pub
• Training Language Models to Explain Their Own Computationsarxiv.org