My main questions: Will these repositories be used in production by Twitter? Is this now the mainline, not a semi-regularly-synced mirror?
Of course not
Twitter's Recommendation Algorithm
191–200 of 1001 posts
Re: Twitter's Recommendation Algorithm
#192From https://github.com/twitter/the-algorithm/blob/7f90d0ca342b92... ( "author_is_elon", candidate => candidate .getOrElse(AuthorIdFeature, None).contains(candidate.getOrElse(DDGStatsElonFeature, 0L))), ( "author_is_power_user", candidate => candidate .getOrElse(AuthorIdFeature, None) .exists(candidate.getOrElse(DDGStatsVitsFeature, Set.empty[Long]).contains)), ( "author_is_democrat", candidate => candidate .getOrEls…
Only used for metrics, apparently. [0] /** * These author ID lists are used purely for metrics collection. We track how often we are * serving Tweets from these authors and how often their tweets are being impressed by users. * This helps us validate in our A/B experimentation platform that we do not ship changes * that negatively impacts one group over others. */ [0]: https://github.com/twitter/the-algorithm/blob/7f…
... Metrics tracked in AB test. So even if it's not explicitly encoded in the algo (or implicitly through some of the features plugged in), they'll pick the winning cell as long as it doesn't hurt Elon's metrics (I'm just parroting the comment you quoted).
It doesn't have to be in the algorithm for the systems to be tweaked to please Elon vanity metrics.
[I've been running lots of ML AB tests over the years, some in organizations of similar size & complexity as Twitter]
Re: Twitter's Recommendation Algorithm
#193Earlier quoted context omitted.
Is really "Following" the entire chronological feed? I feel I miss tweets from people I follow that actually appears in the "For You" tab.
It's not. I follow a fairly small # of people (~500) and getting people to reliably show up in is a long-running problem. Following is not enough, you have to favorite or somehow interact with them sufficiently to be sure of seeing all their tweets. It's quite annoying.
Are you sure? I didn't notice something like this.
Re: Twitter's Recommendation Algorithm
#194If Twitter was 'dead' why on earth are we still talking so much about this blue bird site?
It looks like once again these lot predicting that he won't open source the algorithm and are going to start eating their words again [0], just like they did around incorrectly predicting Twitter's immediate collapse [1] and will look at the source code anyway and continue to talk about "Twitter" again.
If Twitter can open-source their algorithm, Why not TikTok? Either way, the bots are now going to have a very expensive time on Twitter.
Re: Twitter's Recommendation Algorithm
#195They didn't open source the data the censoring abusive, toxicity, and nsfw the algorithms check against, so I'd call it a partial open-sourcing.
Re: Twitter's Recommendation Algorithm
#196Earlier quoted context omitted.
This is 100% not their working copy.
Well yes, but I am pretty sure, elon envisoned warm welcome by the OS community and help for free and now that is off to a bad start.
I doubt he thought he'd get any help for free, given I don't think you can run this code there isn't a meaningful way to contribute.
It's just about transparency or PR, take your pick.
Re: Twitter's Recommendation Algorithm
#197Earlier quoted context omitted.
The author_is_elon flag doesn't surprise me, but the two political designators are somewhat shocking. I'd sure like to know what changes based on what Twitter knows about your political affiliation.
So many questions. How are users tagged D or R? Is that a manual process or automated somehow? What is the effect of these tags? Can I find out if my Twitter account is in one of those buckets?
Especially for people that aren't... you know.. Americans.
Unless they mean actual public figure party members which are known and probably verified.
Re: Twitter's Recommendation Algorithm
#198I wonder what determines 'cred' for this part: https://github.com/twitter/the-algorithm/blob/7f90d0ca342b92...
I answered my own question https://github.com/twitter/the-algorithm/blob/7f90d0ca342b92... "This method reduces the page rank of users who have a low number of followers but a high number of followings."
Heh, I knew it. You need to prune your own following list regularly or become less and less visible. I suspect (but have yet to check) that they also weight visibility in terms of historical follower growth.
That's why you see so many trolls with very low follower counts; it's more effective to make/purchase a new firstname-bunchanumbers account and poop in people's replies than to let Twitter decide placement based on historical factors.
Re: Twitter's Recommendation Algorithm
#199Re: Twitter's Recommendation Algorithm
#200Context: I teach at Princeton and study social media and recommendation systems. From a very quick skim of the repositories, this appears to be quite limited transparency. The documentation gives a decent high-level overview of how Tweet recommendation works—no surprises—and the code tracks that roadmap. Those are meaningful positive steps. But the underlying policies and models are almost entirely missing (there are…
What about these? https://huggingface.co/Twitter