Anthropic, Claude 내부에서 '패닉' 단어 발견
Anthropic는 새로운 기법으로 Claude 내부를 들여다보아 출력에 영향을 미치는 수백만 개의 숨겨진 특징을 발견했으며, 모델이 자신의 내부 상태를 모니터링할 수 있음을 밝혔습니다.
anthropic released a study on 'mechanistic interpretability', mapping millions of internal features within its claude models to decode their decision-making processes. researchers identified a 'j-space' where concepts—like 'panic' or 'protein'—appear internally to guide reasoning without surfacing in the final text. this demonstrates that large language models maintain complex internal states and self-monitoring capabilities, suggesting that future alignment work could target these hidden layers to ensure safety rather than just observing external behavior.