Common Sensitivity: The Future of Knowledge

The larger models do have more of it. So that's really exciting because we might be able to finally give a major crack at it. What's quite curious though is that it randomly fails at some of these simpler cases. That's amazing. But it's also really curious why does it fail at such easy cases that children would be able to answer correctly.