4 ms·
Aren't LLMs still bad with letters inside words and such?
by actionfromafar 3y ago
Aren't LLMs still bad with letters inside words and such?
- crooked-v 3y agoThe point is that any response by any current LLM to those instructions will be obviously distinct from the way a human will respond.
- EForEndeavour 3y agoIt strikes me as an arms race between cute little tests like this one and the ability of LLM app dev teams to mitigate them.