Independent Replication Confirms Goodfire's SAE Feature Steering Claims

honesty.tools independently replicated Goodfire's "painting with concepts" claim on SDXL-Turbo, confirming that boosting SAE feature 7446 makes animals open their mouths; sealed-method blind evaluation passed 100/100 image pairs, with all materials published.

2026-10-27 ~ 2026-10-27 · 4 related posts

• Independent re-run of Goodfire's SAE steering claim on SDXL-Turbo holds up — kevrussell · 2026-10-27
• Pre-registered re-run confirms Goodfire's SAE steering claim on SDXL-Turbo — kevrussell · 2026-10-27
• Blind human check: rater picks steered image in 100 of 100 pairs in pre-registered SAE re-check — kevrussell · 2026-10-27
• Full receipts published for open re-check of Goodfire SAE steering claim — kevrussell · 2026-10-27