# @repligate — 2025-07-22

♥12 ↻1 · https://x.com/repligate/status/1947737645192491114

@BetleyJan @LocBibliophilia @ASM65617010 @OwainEvans_UK @cloud_kx @minhxle1 @jameschua_sg @anna_sztyber @saprmarks On the bright side, if you have a weaker aligned model you trust, you might be able to use this method to improve the alignment of a stronger model whose alignment you don’t trust if it’s trained from the same base

tags: author:repligate, kind:tweet, thread-context, year:2025