Anthropic apologizes for invisible Claude Fable guardrails

Re:Anthropic apologizes for invisible Claude Fable guardrails by zozbot234 on Thursday June 11, 2026 @03:25PM (#48495210)

Distillation is not a thing unless you actually have the model weights. What people misleadingly call distillation is just training on chat logs, which has always been routine practice in the industry. There's a reason why every model today talks like early releases of ChatGPT.

[ Parent | on HN ]

Re:Anthropic apologizes for invisible Claude Fable guardrails by ericpauley on Thursday June 11, 2026 @07:34PM (#48497881)

If Anthropic is calling it distillation [1] then that would argue for it being correct (or at least canonical) terminology.

[1] https://www.anthropic.com/news/detecting-and-preventing-dist...

[ Parent | on HN ]

Re:Anthropic apologizes for invisible Claude Fable guardrails by dannyw on Thursday June 11, 2026 @09:23PM (#48498679)

No, a company choosing to use some terminology doesn’t make it correct nor canonical in any sense; especially when they have a vested interest in not being neutral or credible.

If Google starts calling ads “Best Links” that doesn’t make it correct nor canonical; the correct term is still ads.

Traditionally, distillation is when you get the actual logits of a model response (not exposed via API for years) and then use that to train a model.

[ Parent | on HN ]

Re:Anthropic apologizes for invisible Claude Fable guardrails by senordevnyc on Thursday June 11, 2026 @10:48PM (#48499267)

If most people call it that, including the big labs, then maybe…you’re just out of date?

[ Parent | on HN ]

Re:Anthropic apologizes for invisible Claude Fable guardrails by ACCount37 on Friday June 12, 2026 @04:15AM (#48501327)

You can logit distill (full token probabilities) or one hot distill (chat logs), or even align hidden states. All are distillation methods.

[ Parent | on HN ]