would love to learn more about the latest with deterministic / provable LLMs (is zkML the term?)

LLMs as reproducible compilers, compression algos, embedded in smart contracts, proof that handlers aren’t larping as agents, etc etc

who’s got the inside scoop here & can share some info to enlighten the rest of us?

on unruggable agents:
https://nousresearch.com/setting-your-pet-rock-free (nous x flashbots)
https://phala.network/posts/truth-of-AI-Agent (prominent TEE infra shop)

smol forum chat on the dstack framework used to impl that^ agent:
https://collective.flashbots.net/t/dstack-speedrunning-a-p2p-confidential-vm/3876

https://github.com/zkonduit/ezkl is probably still sota rn

@remco from Worldcoin had a cool talk on this in Devcon from a few years ago and prob best source for practical knowledge on zkML / zk neural nets.

AFAIK for any model reasonably large or useful by modern standards though TEEs are the closest you can get today — cool stuff from Nous research here

Agents acting in TEEs are one. Basically prove what you deployed and that you haven’t tampered with it