
Amanda AskellResearcher
Anthropic
Research Scientist (Alignment Finetuning)
via Anthropic
Researcher · Engineer

Member of Technical Staff, Societal Impacts at Anthropic
Sorensen develops pluralistic alignment: methods intended to help language models represent differing human values rather than collapsing them into one assumed consensus. At Anthropic's Societal Impacts team, he studies how models handle subjective judgments and conflicting perspectives.
See something inaccurate or outdated?
Anthropic
Humans&
University of Washington
Google DeepMindWork and education history is primarily focused on AI-relevant roles and may not be comprehensive.
Last editorial review: August 8, 2026