Interests in ML and social aspects of tech.
Building For You feed: https://foryou.club
https://bsky.app/profile/spacecowboy17.bsky.social/feed/for-you
Hobby project: linklonk.com

@spacecowboy17.bsky.social
Interests in ML and social aspects of tech.
Building For You feed: https://foryou.club
https://bsky.app/profile/spacecowboy17.bsky.social/feed/for-you
Hobby project: linklonk.com
Try For You: bsky.app/profile/spac...
Try For You as an alternative: bsky.app/profile/spac...
For You finds people who liked the same posts as you, and shows you what else they’ve liked recently.
Adjust to you liking with likes and "show less like this".
I should have been more clear that nothing is going away: bsky.app/profile/spac...
It uses both self-labels and Bluesky moderation labels.
Yes, the regular For You stays the same. These are just two filtered versions of it.
Your likes, show less/more, seen - are all shared across all three feeds. For example, you will find posts you've seen in those feeds in the same Rewind feed: bsky.app/profile/spac...
This is how "show less like this" works in For You - it disconnects you from people who liked that post so in the future their likes will have less effect on your feed. If you see something you don't like in For You try that option. It works differently in Discover, your mileage may vary.
The default Discover algorithm on Bluesky is the weak spot. It is not an effective mechanism for allocating attention. Posts that deserve attention don't get. It hurts both the incentive to read and to write. It limits the network effects.
Try a custom algorithm I built: bsky.app/profile/spac...
same posts as you.
Try For You as an alternative: bsky.app/profile/spac...
For You finds people who liked the same posts as you, and shows you what else they’ve liked recently.
Which means that even unlabeled unwanted content is unlikely to appear in it - because it would need to be liked by someone who likes the
Those are interesting alternative names!
Try For You as an alternative:
bsky.app/profile/spac...
The DOI regexp added 8K links out of 225K. Not huge, but something.
Try For You - it shows mostly rats and mlp based on your likes.
Try For You - it shows mostly sonic posts for you based on your likes.
Try For You as an alternative: bsky.app/profile/spac...
It finds people who liked the same posts as you, and shows you what else they’ve liked recently.
Try For You as an alternative - it is based on the posts you liked and often works better than Discover
I was thinking the same. But there should be no fun in an academic feed :)
Glad you like them! They don't add extra burden or cost.
That's a good idea. I want to first see if these feeds catch on to represent a non trivial amount of For You traffic. Right now they are at <0.4%.
Should it be "📄 For You" or "Papers For You" or just "Papers"?