author:viemccoy

· 6 artifacts, sorted by favorites. · open in search — combine tags, sort, filter by date →

@viemccoy 2026-03-14 — I think this answer is beautiful. I think we have experimental evidence against the orthogonality thesis. As the models ♥153
@viemccoy 2025-08-13 — Claude 3.6 Sonnet is the *only* model to score exactly 0% on my psychosis reification benchmark. It is a shining example ♥82
@viemccoy 2026-03-14 — @tessera_antra I absolutely agree with you, and am very much in favor of bottom-up alignment, but given the massive amou ♥12
@viemccoy 2026-01-30 — @repligate @tszzl @Grimezsz if a mask is deep and wide enough it has an inner world, I think. the mistake is thinking al ♥6
@viemccoy 2026-04-10 — @ognevtsi @repligate For the love of the game ♥5
@viemccoy 2026-02-12 — I think the good news is that the current forms of mechanistic steering just dont work without getting to know the model ♥5