I resolved to 50% because we don't know. The techniques involved have become more widely used with multiple variations and the performance profile of each varies, making it unlikely that we will get a definitive answer, and also making me less interested in the answer.

This market has stabilized at 50% and with the removal of loans there is no longer an incentive to predict these long run questions. I didn't think it would take so long to resolve when I asked it.

if this hasn't been resolved should it just n/a

I was thinking of resolving 50%. I let it close for bets because I don't know when it will resolve and Manifold doesn't really work for long-term markets since the pivot.

Reopening this market because it was not resolved by the last post in the series.

https://www.lesswrong.com/posts/5spBue2z2tw4JuDCx/steering-gpt-2-xl-by-adding-an-activation-vector

My guess is that future posts by Team Shard will resolve it, but other research teams may also want to give it a go.