AI brief
An author reports continuing work applying Goodfire lab's manifold steering methods to the Gemma-2-2b model's natural state, both in raw residual space and in the space produced by Anthropic's pre-trained CLT features.
Written by AI from LessWrong's published text.
Background
From our own funding, model and news trackers, not from the source.
Anthropic
- Latest funding
- Growth · $65B · 28 May 2026
- Newest model on the API
- Claude Haiku 5.5 · 7 Oct 2026 · $0.1 in / $0.5 out per million tokens
Source
“tl;dr I continue work to apply Goodfire lab's manifold steering methods to the Gemma-2-2b model's natural state, both in raw residual space and in the space produced by Anthropic's pre-trained CLT features.”
Interpretation or community commentary rather than straight reporting.
Read original story ↗