[QuantumGood]

Just to clarify your implication: Fable subscription usage was just (re)extended to July 19

[maxgashkov]

If they haven't relaxed the classifier this makes little to no difference as the model is essentially unusable no matter the token quotas.

[anuramat]

what are you working on? I only hit the guardrails twice after burning through two weeks of 20x max plan, both times on ML stuff; still more than I'd want to, but not unusable

[tomp]

I was downgraded to Opus after asking “how do fish get into mountain lakes”. Fable managed to get like half an answer through…

[vardalab]

A lot of stuff that has to do with VLLM and troubleshooting and compiling and building VLLM, compiling kernel, or just dealing with setting up eval for local models, it punts to Opus 4.8 on a regular basis. To the point that I have given up on using it for that purpose.

[dannyw]

Hey, at least you’re not being silently nerfed or sabotaged by Fable’s PEFTs or steering vectors instead. Those classifiers are only meant to target DeepSeek et al, not routine LLM or ML work /s.

[fragmede]

Lucky for you. I, along with everyone else on the fable guardrail thread have been hitting guardrails with Fable for boring normal shit and getting downgraded.

[usef-]

Interesting how different experiences are. I haven't hit any guardrails after a lot of dev the past week. (but I'm working on motion gesture code and similar things)

[anuramat]

what languages are you working with? I imagine if it's something like C, you'd hit guardrails every time you manage memory

[maxgashkov]

very mundane web app and networking security (vpn etc.), nowhere near redteaming

[anuramat]

> nowhere near readteaming

I think any security-related task triggers it to think about the threat model and thus hit the guardrails

[datadrivenangel]

And now their comment is good next week! The rug pull getting extended doesn't mean it won't come.

[QuantumGood]

Their comment was good as is because of the extension (not a nitpick, a clarification)