Remix.run Logo
0x696C6961 3 hours ago

It would be pretty simple to see what API they're calling.

consumer451 3 hours ago | parent [-]

That's what I meant to get at by "it runs on their cloud."

They can name that user-facing ultrareview API endpoint whatever they want, and we have no way to see what model endpoint it calls internally once running on their cloud, right?

zarzavat 2 hours ago | parent [-]

Introduce intentional and increasingly subtle vulns and test against Sonnet, Opus, etc? Should give statistical evidence of its power.