extremely cool. I’ve had Claude walk model architectures to debug Loras and fine tunes, but this is delightful

I'd love to hear more about that. Are you able to fine tune specific layers?

Specifically, one particular otherwise excellent model I use has an alignment problem (sycophancy) that I've isolated to a specific layer. I can nuke the layer with lora and the behaviour stops - but I'm not sure what else I'm nuking in the process. I'm quite new at this so I'd love any advice. Thank you!

Their enterprise customers pay via metered actual use.

making sure ML isn’t ever pigeonholed

Incredibly thoughtful. This essay gives that very rare sense of being well reasoned, gods at forest and trees, and sitting atop a shit ton of domain expertise.

As an Aubrey-Maturin lifelong re-reader, I cannot wait to play this.

They have all switched to usage plus cheap seats based costs for enterprise contracts. the seat costs are typically 20-35% of total spend.

iirc, there is a bunch of formal machinery you need to define probability distributions for situations such as infinite outcomes (eg what is the probability that a random real number between 0 and 10 is less than 3?)