X Shares New Insights Into Transparency and Shadowbanning
The release includes model settings, filters and ranking parameters, and a pilot tool will let some users download account labels as JSON.
- X released the source code for its 'For You' timeline on GitHub under the Apache 2.0 license on Thursday, alongside a new 'Under the Hood' tool letting users review whether their accounts or posts have been affected by ranking systems.
- Elon Musk previously vowed to fully open-source the codebase to win user trust; however, past efforts were criticized as partial, leaving many users frustrated by the platform's perceived lack of transparency.
- Users who have posted ten or more times over the past month can download aggregate JSON stats via 'Under the Hood' to identify visibility-limiting labels including NSFW and Civic Integrity flags.
- X's VP of Product Keith Coleman called the changes 'an unprecedented level of transparency into the X algorithm,' enabling users to audit the ranking system by processing their data through an LLM.
- While the company claims this addresses algorithmic concerns, it continues to withhold code related to ads and Grok, citing risks that bad actors could exploit the system to evade platform rules.
14 Articles
14 Articles
X reveals the formula for going viral on its platform
X unveiled the source code for the algorithm that determines what shows up on its For You page.Illustration by Thomas Fuller/SOPA Images/LightRocket via Getty ImagesX on Thursday unveiled the source code for its For You page algorithm.The code contains positive and negative "weights" based on user reactions to posts.A new feature also allows users to see labels applied to their accounts.On X, not all engagement is equal.Elon Musk's social media …
X open-sources its ranking algorithm, lets users check account-level impacts
X is expanding access to its recommendation-system code and introducing new transparency tools for users, while withholding some algorithmic details that it says could help bad actors evade its rules.
X open sources its ranking algorithm, letting users see if they've been 'shadowbanned'
X is expanding the open source code behind its 'For You' feed and launching new transparency tools that show users when its ranking systems have affected their accounts or posts.
Coverage Details
Bias Distribution
- 34% of the sources lean Left, 33% of the sources are Center, 33% of the sources lean Right
Factuality
To view factuality data please Upgrade to Premium
