Varun Srinivasan on Warpcast

Varun Srinivasan pfp

Varun Srinivasan

AMA - if you have any questions about how spam detection works in Warpcast, I'll answer them over the next few hours.

111 replies

89 recasts

269 reactions

hyperbad pfp

why am i averaging 0.5 likes per post when they are bangers ಠ_ಠ

1 reply

0 recast

4 reactions

Nicola Ceornea - ethOS pfp

Nicola Ceornea - ethOS

How can someone see if they have been labeled as spammers? Will people that have been already labeled as spammers received the notification?

2 replies

0 recast

3 reactions

Joe Hathaway pfp

I’m just curious about how to define the extent of “excessively”? Would this algo be friendly to new normal users on this platform? https://warpcast.com/dwr.eth/0x32f99aea

2 replies

0 recast

1 reaction

Daisy✨ pfp

How does Warpcast handle false positives in spam detection to ensure legitimate users aren't unfairly flagged?

1 reply

0 recast

1 reaction

Maretus pfp

Is the current spam system based on reports from other users, algorithmically, or a combination of both? If it’s a combination, are they weighted equally? Ie - does a user spam report carry more/less weight than an algorithmic determination?

1 reply

0 recast

1 reaction

Commstark 🎩🫂 pfp

Commstark 🎩🫂

is the following correct: there is an overlap between the definition of Spam and new users being deprioritized in larger account feeds e.g. if you're new and post on a larger account, the larger account can't see you so they can't reply. This drastically increases the odds of you being labelled spam

1 reply

0 recast

0 reaction

Dvyne🎩 pfp

If it's by interacting with people you follow to avoid being labeled as spam... how do you want new users to interact then and not get labeled? Cause they're practically new, no one is following them.

1 reply

0 recast

0 reaction

ihawaii.eth🎩🎭 pfp

ihawaii.eth🎩🎭

would you consider a vouching system for friends who were wrongfully marked as spam? for example, having a certain number of OGs and big names in the community verify and vouch that users aren’t spammers. would you then clear their spam label manually?

2 replies

0 recast

0 reaction

Mikado pfp

Do auto tip distribution systems like /microsub and /impact have a negative effect on my account?

1 reply

0 recast

1 reaction

Dara Khan pfp

@spookyaction-btc

Have you guys considered rewarding users who take the time to identify spam/bots and do a good job of it? Crazy to me that X hasn't done this.

1 reply

0 recast

0 reaction

Mac Budkowski ᵏ pfp

Mac Budkowski ᵏ

Whats the most non-obvious thing that makes you labeled as spam?

1 reply

0 recast

0 reaction

⋆♱ 9̷0̷†մղ ♱⋆ pfp

⋆♱ 9̷0̷†մղ ♱⋆

I’ve noticed that when opening a new account, for accounts that follow other accounts within the first few hours, they all go into spam. Is activity when you just open an account tied to spam? @v

1 reply

0 recast

0 reaction

Alemac pfp

The spam model reclassifies users automatically? Or it is done by the team?

1 reply

0 recast

0 reaction

zoo pfp

tangential q. sorry if this has been asked before. how does the 'Show fewer like this 👎' action play into the algo and spam detection?

1 reply

0 recast

0 reaction

zenxosa🎩 pfp

If you get labeled as spam and you create another account,do you have higher chances of falling into spam?

1 reply

0 recast

0 reaction

Jer | seattlecrypto.eth pfp

Jer | seattlecrypto.eth

@seattlecrypto.eth

What’s the metrics in place that determines if you get the label? Is it a percentage of your replies to people who don’t follow you? A certain amount of replies within a short amount of time?

1 reply

0 recast

0 reaction

law pfp

2. I know you said we should ask how spam detection works. But this is as important. Can we have a channel or maybe a few people, like a support team, who we can tag to things like new users being wrongfully marked as spam, who can help to take necessary actions. It’s always gonna be harder to get your attention or Dan’s.

1 reply

0 recast

0 reaction

Harry pfp

1. How often does the spam detection detect a spam users and apply the label? 2. If users marked as spam, is there a chance for them to unlabel it?

1 reply

0 recast

0 reaction

law pfp

Okay i have a couple of questions In the explanation of spam detection or what could trigger it… We have only heard about engaging people who don’t engage back Are there no other means by which one could risk being put in the spam list like the excessive use of words repeatedly, posting only pictures for repeatedly and so on.

1 reply

0 recast

0 reaction