Paper Emergent Introspective Awareness in Large Language Models Anthropic · 2025 Read the paper → introspectionmech-interp
