SafePlanBench (work in progress)

[Agustín Martinez Suñé]

Hi all,

I recently gave a talk at the Guaranteed Safe AI Seminars: Using PDDL Planning to Ensure Safety in LLM-based Agents.

I’d like to bring your attention to our Manifund post: SafePlanBench: evaluating a Guaranteed Safe AI Approach for LLM-based Agents. We’re raising funds to support the development of a benchmark that systematically evaluates the safe planning capabilities of LLM-based agents. Any support, whether through donations or sharing the post, is appreciated!

I hope it’s okay to share this here.

Best,

Agustín.

[Evan Miyazono]

Just personally contributed the first $500 - good luck with the rest!

--
You received this message because you are subscribed to the Google Groups "guaranteed-safe-ai" group.
To unsubscribe from this group and stop receiving emails from it, send an email to guaranteed-safe-ai+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/guaranteed-safe-ai/CALoAswHAV3JGv4T9qBOgKGgEEdKfAnR1qzN_LFFDWUG4mZJ4yg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.