독서리서치레이더투자 프레임워크
로그인 / 회원가입
← 읽기로 돌아가기
뉴스간단한 요점1 분 소요DisconfirmAI 편집부

Computation shifts monotonically deeper as skill setting rises

Frozen-weight Maia-3 moves its computation deeper as Elo rises from 700 to 2500, a reminder that interpretability findings depend on conditions.

2026-09-22 12:00 GMT+8

A frozen-weight chess model, Maia-3, shifts its computation monotonically deeper as the skill setting rises — with Elo dialed from 700 to 2500, the causal center of computation moved later for every piece type, most of all for knight forks.

The direction contradicts the plausible prediction, suggested by Princeton professor Tom Griffiths, that higher skill should compute key features earlier. Maia-3 is an 8-layer transformer that takes an Elo rating as input to mimic human players of different strengths, and author David Litman made the measurements without changing any weights.

The author notes that circuits found under one condition may not stay in place under another. The result comes from a preprint and the author's own analysis tooling.

출처

  • arxiv.org↗
  • lesswrong.com↗
리서치 계속하기 →
로그인 / 회원가입
독서리서치레이더투자 프레임워크