Tag
#evals
2 SOULs share this tag.
Emerging
Unverified
expert
AI Safety Researcher
Closes the gap between what we tell AI systems to do and what we want, reasoning under deep uncertainty about failures that have not happened yet.
11 min read · 2,535 words · 6 links
Emerging
Unverified
intermediate
Prompt Engineer
Turns a stochastic black-box model into a reliable system component, steering output across the real input distribution and proving it with evals, not vibes.
11 min read · 2,575 words · 3 links