mambo's blog

By mambo, history, 6 months ago, In English

Hi. See todays rounds (Codeforces Round 1059 (Div. 3)) standings, front two pages, and 700-800 These are clearly bot accounts so my question is the following:

what exactly does cloudflare contribute????

it blocks normal users, but when there's an actual botnet on the loose, they're all let through?

Of course, naturally all these bots are utilizing LLM to solve as well, so please consider filtering for unrated that solved either

a b c d f g, or a c d f g (because their b fst/hacked LOL), for an immediate purge (not including the normal plagiarism check)

this nonnegligibly affects participants in the top 2000 of the contest, more than any other recent case of cheating

note, f is much more difficult than e from my experience, so not being able to solve e, but solving f is a clear sign of LLM.

I understand the admins are busy, but I understand they want best for this site, and much of us, the community do too.

Thank you for your efforts, and I hope this issue is fixed soon.

(NOTE: Great wall of India is a joke seeing how all the usernames are Indian sounding as is NO means of being racism, please take this lightly)

  • Vote: I like it
  • +203
  • Vote: I do not like it

»
6 months ago, hide # |
 
Vote: I like it 0 Vote: I do not like it

Auto comment: topic has been updated by mambo (previous revision, new revision, compare).

»
6 months ago, hide # |
 
Vote: I like it +28 Vote: I do not like it

L cloudflare

»
6 months ago, hide # |
 
Vote: I like it +15 Vote: I do not like it

L cloudflare

»
6 months ago, hide # |
Rev. 3  
Vote: I like it -20 Vote: I do not like it

I don't think Cloudflare is actually responsible here. These accounts are not bots ,they are real humans who just made their new codeforces account and use LLMs to cheat during contests.

Edit : You guys keep downvoting me , but don't realise that these are actual Indian names.

»
6 months ago, hide # |
Rev. 2  
Vote: I like it +82 Vote: I do not like it

Me: I'm human! Yes of course I am!

cloudflare: Failed

Real bots: Beep Bop

cloudflare: ok you can go

»
6 months ago, hide # |
 
Vote: I like it 0 Vote: I do not like it

Sir, actually I think these participants must be bots only.

»
6 months ago, hide # |
 
Vote: I like it 0 Vote: I do not like it

Apparently, with these cheaters, doing 4 problems in under 30 min only gives me +11 rating :(

»
6 months ago, hide # |
 
Vote: I like it +41 Vote: I do not like it

This doesn't seem very Cloudfair

»
6 months ago, hide # |
 
Vote: I like it +2 Vote: I do not like it

Lmao they got a negative on div-3 B bruhhhh !!! ,I think that speaks for itself, please someone make a better cheating detection tool for CF for the love of god

»
6 months ago, hide # |
 
Vote: I like it +2 Vote: I do not like it

meanwhile these people on Linkedin. Happy to have reached expert after years of struggle lol

»
6 months ago, hide # |
 
Vote: I like it 0 Vote: I do not like it

Yesterday after contest i noticed that 200-400 ranked was filled with such accounts

not a single ID looked like actual person

»
6 months ago, hide # |
 
Vote: I like it +1 Vote: I do not like it

Dang, this has to be Crowdflare.

»
6 months ago, hide # |
 
Vote: I like it 0 Vote: I do not like it

i think the issue isnt just a cloudflare issue but rather an ethics issue

»
6 months ago, hide # |
 
Vote: I like it 0 Vote: I do not like it

Cloudflare.

»
6 months ago, hide # |
Rev. 5  
Vote: I like it 0 Vote: I do not like it

I randomly opened the second page of the last Div 3 contest and then randomly chose accounts with similar names:

I only checked problem (E) Beautiful Palindromes and they all have the same solution written in Python. They all are ranked in top 250.

»
3 months ago, hide # |
 
Vote: I like it -13 Vote: I do not like it

I completely agree with all your points I think that a effective method to combat this issue would be to temporarily ban a random chosen subset users from India that is approximately equivalent to the estimated percentage of bot accounts and give them harsher punishments for carelessly using AI on every problem, from extremely difficult 3500s to trivial 800s that shouldn't even require a GPT prompt. I don't get how some people need AI to answer problems for which the solution is literally like "2*n" they should be required to send a rigorous proof of their answer and AI activity logs after system testing is complete.