AI - AI Dynamics

Task Emergent for Different Metrics Than Initially Measured

By

–

03 May 2023 20h58

sure, the task would not be emergent *for that metric*, but its still emergent *for the metric we care about*

→ View original post on X — @midjourney

Midjourney Magazine Issue 2 Coming Soon After V5 Release

By

@midjourney

–

03 May 2023 20h41

Midjourney Magazine Issue #2 will be going to the printing presses soon! If you want a copy, please make sure you're subscribed by Friday at http://
mag.midjourney.com. Issue #2 will be dedicated to images made in the last ~45 days since the V5 release.

Discussion on Different Metrics in Previous Research Paper

By

–

03 May 2023 20h23

Finally, some discussion around using different metrics was given in our previous paper, which could be worth taking a look at: https://
openreview.net/pdf?id=yzkSU5z
dwD
…

Discussing Emergent Abilities in AI: Risks and Predictions

By

–

03 May 2023 20h23

Overall, I’m glad that the idea of emergent abilities is being discussed more. I’m excited about work that would enable us to predict emergent behavior, since emergence includes risks as well as abilities. I’d love to discuss more with you on twitter or at the next conference!

Chain-of-Thought Prompting: Emergence in Large Language Models

By

–

03 May 2023 20h23

Additional response 2: Another popular example of emergence which also underscores qualitative changes in the model is chain-of-thought prompting, for which performance is worse than answering directly for small models, but much better than answering directly for large models.

U-shaped Performance: Emergent Abilities in Model Scaling

By

–

03 May 2023 20h23

Additional response 1: Many emergent abilities also cannot be explained by these arguments. Consider the below plots which show a U-shaped phenomena: performance actually decreases for several model scales, until it suddenly spikes up again.