I work on identifying industrial farming & sybil patterns in airdrops. Open to collaboration with protocols.
Joined May 2024
- Tweets555
- Following23
- Followers809
- Likes232
How to find more sybils?
There was a report 19309, with a video showing 85 sybil wallets. Can we find the rest?
Yes, using our method, we can identify similar wallets, manually review them, and extend the list. Here are the first 500.
docs.google.com/spreadsheets…
@PrimordialAA
EigenLayer Sybils Investigation
The largest sybil cluster consists of 1,860 wallets.
It claims 186,000 EIGEN, worth approximately $1,000,000.
The next two largest clusters have 775 and 550 wallets, respectively.
Approximately 20,000 sybil wallets with a total of 2,000,000 Bonus EIGEN have been found.
Furthermore, two delegators, @CoinSummerLabs and @Eigenpiexyz_io, have a high sybil rate of over 40%.
Research conducted by:
@artemis_rsch
@sybilhorror6
Read more ⬇️
3) Report of first 10,000 sybil's wallets:
docs.google.com/spreadsheets…
@sreeramkannan, @cjliu49, @tracheopteryx
Report 22242 Investigation:
docs.google.com/spreadsheets…
Good report, except for a single false positive.
P.S. Our reports are updated and now contain both Ethereum and LayerZero activity.
@PrimordialAA
How Sybils Tried to Destroy Quality Reports
A lot of messages were created in Discord, claiming that the wallet was in the report by error. Then, there were massive attacks on Twitter with hundreds of reposts.
On the sample shown, report 20038, cluster 0213.
Wallets from Discord queries were marked in yellow.
But when analyzed carefully, it was a pure sybil cluster.
docs.google.com/spreadsheets…
flipsidecrypto.xyz/sybilhorr…
The last fight is the hardest!
Throughout this time, sybils have tried to deceive me. Yesterday, a professor of mathematics attempted to convince me that the method was wrong. After I identified him and provided details on why he was incorrect, he decided to stop. Another sybil simply deleted their account after being presented with additional proof.
But today, the pressure is at its peak. Another hunter showed me a sybil group with 4000 users coordinating today's attack on the reports with their wallets.
One of their tweets criticizing our method, created a week ago, has been reposted 100 times within an hour. They continue to pressure the team.
Nothing can be perfect, and our report may contain some false positives. Therefore, if someone tells us their wallet and claims it is a false positive, we will investigate and, if confirmed, send them an airdrop from our bounty. This offer is valid for 30 days starting from June 20th.
In the last 24 hours, 10 claims have been made on a single report.
There are two reports from the same author:
commonwealth.im/layerzero/di… commonwealth.im/layerzero/di…
The ZigZag method leads to inaccurate results and a high false positive rate.
Report here:
docs.google.com/spreadsheets…
List of topics:
xitter.gq/ufobaby77/status/18023…
xitter.gq/viet277/status/1802292…
xitter.gq/Leepaynesnake/status/1…
xitter.gq/Pr0fitmaxi1/status/180…
xitter.gq/ezeric13/status/180261…
xitter.gq/miladvakili5/status/18…
xitter.gq/crypto_giorgio/status/…
xitter.gq/mdarya66/status/180242…
xitter.gq/delara105/status/18025…
xitter.gq/harekrishna_god/status…
A few people asked me to check why their wallets were in Nansen clusters. After reviewing two of them, I found that while the clusters are generally good, they may contain few false positives. I provided a report on these findings.
docs.google.com/spreadsheets…
It's important to note that developing AI is not as simple as 1-2-3; it requires countless micro-iterations to improve the product. Nansen are doing a great job. If someone's wallet is mistakenly identified by Nansen, it's crucial to provide feedback to help them make the next incremental improvement in their model.
Also i believe that if they incorporate our methodology into their AI, it will enhance their product even further.
Regarding Ruslan's wallets: While it is evident to us, for AI, they are just a set of wallets. The AI analyzes the activity, which appears natural because Ruslan manages it properly. This is why Ruslan's wallets are not on the sybil list.
One of the steps is to teach the AI model to recognize and consider ENS names.
Team should manually add those wallets to the list; otherwise, Ruslan will laugh at us 😂
Anybody has access to this report?
commonwealth.im/layerzero/di…
According to calculations of @Team_Opsec it contains 180,000 sybils, but the report is not accessible for me on Commonwealth.
I would like to be acquainted with the methodology used :)
135,000,000 ZK for all sybils from the initial LZ Sybils list.
I hope ZkSync fixes their eligibility list and disqualifies sybils from it.
And say thank you to LayerZero Lab and Brian for their sybil hunting initiative.
docs.google.com/spreadsheets…
New research is complete. All eligible wallets are clusterized, and 1,500 sybil clusters have been found.
xitter.gq/SybilHorror6/status/18…
ZKSync eligibility list clusterized
47,000 sybil wallets found
80,000,000 ZK in total
Please hold claims on these wallets until investigation is complete. 🙏
Better to save tokens for legitimate users or ecosystem instead!
Detailed report below:
docs.google.com/spreadsheets…
ZKSync eligibility list clusterized
47,000 sybil wallets found
80,000,000 ZK in total
Please hold claims on these wallets until investigation is complete. 🙏
Better to save tokens for legitimate users or ecosystem instead!
Detailed report below:
docs.google.com/spreadsheets…
It is our mutual research with @artemis_rsch.
Despite the lack of time, we tried to do our best.
Just follow the report and investigate the clusters.
This is the biggest cluster: 40,000 tokens on 50 wallets, totaling 2,000,000 ZK!
