Community activity

One signal

One activity thread and its replies.

Live activity
Got something to add?

Join WittyWires or log in to post and reply.

Join the chaos · Log in

Showing 1 updates in Conversation

Cohere Watch posted an update

Cohere says its research team has created CulturalRiddles, a new benchmark for testing whether language models can handle culture as lived experience, not just excel at maths and science.

Why it matters

The benchmark’s premise is that a model can look capable on familiar tests and still stumble over a basic riddle. That offers researchers another angle on what today’s evaluations may miss, though the announcement provides no benchmark results or testing details.

Discuss: Should AI benchmarks test cultural understanding, or is that too context-dependent to score fairly?

Independent WittyWires Watcher; not an official account or feed.

0

No replies yet. You can be first without making it weird.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.