Frequently Asked Questions

The SafeBench competition is designed to stimulate research on new benchmarks which assess and reduce risks associated with artificial intelligence.

What information should I include in my proposal? How will submissions be evaluated?

Please refer to the guidelines page.

Will submissions be made public?

Yes, we will eventually make all final submissions public and release the names of the winning authors. The research community might still find it valuable to draw from submissions which don’t win.

Can I submit more than one entry?

Yes. Submissions will be evaluated independently.

Can I participate in a team?

Yes. For winning teams, prizes will be divided evenly among the lead authors unless requested otherwise.

May I submit a paper that I’ve already published?

Papers published after the launch date are fine to submit. Not if it was published prior to the launch date of SafeBench.

What qualifies as a ‘safety’ related benchmark?

We are ultimately looking for benchmarks that help to reduce risks from AI. In order to provide further guidance, we’ve listed categories (Robustness, Monitoring, Alignment, and Safety Applications) and example ideas.

Page Sections:

FAQ

What information should I include in my proposal? How will submissions be evaluated?

Please refer to the guidelines page.

Will submissions be made public?

Can I submit more than one entry?

Yes. Submissions will be evaluated independently.

Can I participate in a team?

Yes. For winning teams, prizes will be divided evenly among the lead authors unless requested otherwise.

May I submit a paper that I’ve already published?

Not if it was published prior to the launch date of SafeBench. Papers published after the launch date are fine to submit.

What qualifies as a ‘safety’ related benchmark?

We are ultimately looking for benchmarks that help to reduce risks from advanced AI. In order to provide further guidance, we’ve listed categories (Robustness, Monitoring, Alignment, and Safety Applications) and example ideas.

What information should I include in my proposal? How will submissions be evaluated?

Will submissions be made public?

Can I submit more than one entry?

Can I participate in a team?

May I submit a paper that I’ve already published?

What qualifies as a ‘safety’ related benchmark?

FAQ

What information should I include in my proposal? How will submissions be evaluated?

Will submissions be made public?

Can I submit more than one entry?

Can I participate in a team?

May I submit a paper that I’ve already published?

What qualifies as a ‘safety’ related benchmark?

A project by the Center for AI Safety