No, it is not an April fool’s joke. Twitter has revealed its source code, as its owner Elon Musk promised. Although it is only a small chunk of Twitter’s source code, it gives us a sneak peek of how the recommendation algorithm for tweets works.
On GitHub, Twitter has shared the source code for its recommendation algorithm. During a Twitter Space conversation about this development, Musk said he hopes users will identify any potential “problems” within the code and contribute to its improvement.
“Our initial release of the so-called algorithm is going to be quite embarrassing and people are gonna find a lot of mistakes but we’re going to fix them very quickly,” Musk said during the Spaces.
The code made public pertains solely to how tweets show up in Twitter’s “For You” feed. However, Musk said that in the coming weeks, more parts of the source code would be released publicly, basically everything that contributes to Twitter’s recommendation algorithm. And Musk also plans to open source Twitter’s search algorithm.
This is how the “For You” feed works
Twitter’s engineering blog revealed that their recommendation pipeline runs around five billion times daily, attempting to select the top 1,500 tweets from a pool of hundreds of millions.
The “For You” timeline currently includes 50% tweets from users not followed and 50% from followed users, though this can vary by user. The ranking process is executed using a neural network with approximately 48 million parameters continuously trained on tweet interactions to maximise positive engagement, such as likes, retweets, and replies.
Twitter does seem to have a “VIP” list
There has been talk about Twitter giving some high-profile users VIP treatment, boosting their tweets, including that of Elon Musk. There are some hints to it in Twitter’s algorithm, which appears to differentiate between tweets based on their source, as researcher Jane Manchun Wong observed that tweets from Elon Musk and those from “power users,” Republicans, and Democrats were labelled differently.
During the Spaces session, a Twitter engineer stated that these labels were only used for metrics, but Musk, who said he was unaware of them, objected to their presence.
Several users have complained of seeing more of Musk on their feeds lately, even though they have never interacted with any of Musk’s tweets.
Several users have complained of seeing more of Musk on their feeds lately, even though they have never interacted with any of Musk’s tweets. Reportedly, Musk ordered the engineering team to push his tweet back in February as he suspected his reach to be declining.
On GitHub, Twitter has shared the source code for its recommendation algorithm. During a Twitter Space conversation about this development, Musk said he hopes users will identify any potential “problems” within the code and contribute to its improvement.
“Our initial release of the so-called algorithm is going to be quite embarrassing and people are gonna find a lot of mistakes but we’re going to fix them very quickly,” Musk said during the Spaces.
The code made public pertains solely to how tweets show up in Twitter’s “For You” feed. However, Musk said that in the coming weeks, more parts of the source code would be released publicly, basically everything that contributes to Twitter’s recommendation algorithm. And Musk also plans to open source Twitter’s search algorithm.
This is how the “For You” feed works
Twitter’s engineering blog revealed that their recommendation pipeline runs around five billion times daily, attempting to select the top 1,500 tweets from a pool of hundreds of millions.
The “For You” timeline currently includes 50% tweets from users not followed and 50% from followed users, though this can vary by user. The ranking process is executed using a neural network with approximately 48 million parameters continuously trained on tweet interactions to maximise positive engagement, such as likes, retweets, and replies.
Twitter does seem to have a “VIP” list
There has been talk about Twitter giving some high-profile users VIP treatment, boosting their tweets, including that of Elon Musk. There are some hints to it in Twitter’s algorithm, which appears to differentiate between tweets based on their source, as researcher Jane Manchun Wong observed that tweets from Elon Musk and those from “power users,” Republicans, and Democrats were labelled differently.
During the Spaces session, a Twitter engineer stated that these labels were only used for metrics, but Musk, who said he was unaware of them, objected to their presence.
Several users have complained of seeing more of Musk on their feeds lately, even though they have never interacted with any of Musk’s tweets.
Several users have complained of seeing more of Musk on their feeds lately, even though they have never interacted with any of Musk’s tweets. Reportedly, Musk ordered the engineering team to push his tweet back in February as he suspected his reach to be declining.
Related Posts
Denial of responsibility! TechnoCodex is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – [email protected]. The content will be deleted within 24 hours.
