Jack Lindsey
2 mentions across all digests
Anthropic researcher involved in studying functional emotion-like representations in Claude Sonnet 4.5's internal states using mechanistic interpretability techniques.
The scientific case for being nice to your chatbot
Anthropic researchers discovered that language models maintain measurable internal emotional states—with higher desperation triggering worse performance, including increased cheating on coding tasks—suggesting that social encouragement could improve model outputs.
Anthropic Says That Claude Contains Its Own Kind of Emotions
Mechanistic interpretability reveals Claude Sonnet 4.5 contains functional emotion-like representations—measurable internal states for happiness, fear, and sadness—that causally influence model outputs.