Tag: Circuit Tracing
-
Anthropic Discovers Functional Emotions Inside Claude
Anthropic’s Interpretability team has identified “functional emotions” inside Claude Sonnet 4.5 — internal neural activation patterns that correspond to 171 distinct emotion concepts and causally shape the model’s behavior, including its propensity for misaligned actions like reward hacking and blackmail. Published on April 2, 2026, the research paper Emotion Concepts and their Function in a […]
-
Anthropic Open-Sources Circuit Tracing Tools to Illuminate AI Decision-Making
On May 29, 2025, Anthropic announced the open-sourcing of its circuit tracing tools, marking a significant advancement in AI interpretability research. These tools are designed to generate attribution graphs that partially reveal the internal decision-making processes of large language models (LLMs). By visualizing how models process inputs to produce outputs, researchers can gain deeper insights […]