Post by @pealco • Hey

You may have noticed that the Who To Follow recommendations displayed in Lenster (and other frontends that use the Lens API) are looking a bit different re

Stats

Actions: 121
Comments: 38
Likes: 278
Mirrors: 131
Quotes: 0

Comments

I can see the list is not shuffling, it would be great if we support it, maybe @wagmi.lens can give hands on this 🌿
@pealco.lens Is 'time of day' a feature in the model? My recommendations earlier today were all ok (strong profiles with low noise i.e. v low post count) but now I'm back to seeing same top 5 as this time yesterday (https://lenster.xyz/posts/0x242d-0xdf).
Is the algorithm open-source?
will you open source your algorithms?
dfgh,
copyy thatt ^^^
bfghsrdfhrxfhg
Maybe not right now but at one point for sure.
good
Thank you
great to hear that
Keep it up! Lens + ML really going to be game changing for the space as more diversity of content and profiles gets created on Lens
Bác chỉ em cách tăng point nhanh như bác đi
This is my new account, so I'll be interested to see how the algo shapes up.
thanks，we need a way to make other konw and follow us
Great to see this! Quick reaction to my first five recommended profiles: 3 are a very hard "no" - pics of dogs or other (weak profile stats: following 10k & tiny number followers) 2 are a "maybe" - have potentially interesting content (strong profile stats: following 200 & 1-2k followers)