Siirry sisältöön
Global sound
LessWrong (30+ Karma)

LessWrong (30+ Karma)

“Lessons From Red-teaming Auto-review, OpenAI’s Blocking Monitor” by monika_j, Giovanni M. Occhipinti

· 29 min

Jakson kuvaus

This work was done by Giovanni Occhipinti, as part of the ERA fellowship and mentored by Monika Jotautaitė (Apollo Research).
Summary
We red-teamed synchronous control measures in Codex, with a focus on its blocking monitor Auto-review. Our campaign directly analyzed the Codex codebase, searching for exploits that are enabled by the scaffold itself, both against the models used as monitors and other measures such as the Codex sandbox, protections, and deny rules.
We found 11 attacks that a misaligned agent could use to defeat the Codex sandbox and Auto-review, executing arbitrary malicious…

Sivusto: lesswrong.com

Livenä, viimeisin tunti: 10 871 kappaletta 4 736 asemalla 133 maassa

Löydä maailma radion kautta

Lisää tapoja löytää (7)

Kuulitko jotain hyvää? Tallenna se tai avaa se suoraan Spotifyssa.

★ Artist of the week The fastest riser on the radio this week
Your song here? From € 9 per week →
"Online Radio (Armenia)"Jatka kuuntelua