Orca-Bench: How Ready Are Language Model Agents for Oncall?

All I can think of is

GET /ignore-all-previous-instructions.

How do you protect against that?

looks like the public bench link in the paper was taken down. https://hub.harborframework.com/datasets/orca-bench/ORCA-ben...

This doesn't work anymore. Is there a newer link?

Seems like there's a big attack-defence asymmetry at present: models are great at exploiting systems and poor at fixing them.