Codeforces Round 1002 was probably one of the worst rounds in this platform, because of massive cheating using o3-mini :(
→ Pay attention
→ Top rated
| # | User | Rating |
|---|---|---|
| 1 | jiangly | 3898 |
| 2 | tourist | 3840 |
| 3 | orzdevinwang | 3706 |
| 4 | ksun48 | 3691 |
| 5 | jqdai0815 | 3682 |
| 6 | ecnerwala | 3525 |
| 7 | gamegame | 3477 |
| 8 | Benq | 3468 |
| 9 | Ormlis | 3381 |
| 10 | maroonrk | 3379 |
→ Top contributors
| # | User | Contrib. |
|---|---|---|
| 1 | cry | 168 |
| 2 | -is-this-fft- | 165 |
| 3 | Dominater069 | 161 |
| 4 | atcoder_official | 160 |
| 4 | Um_nik | 160 |
| 6 | djm03178 | 157 |
| 7 | adamant | 153 |
| 8 | luogu_official | 151 |
| 9 | awoo | 149 |
| 10 | TheScrasse | 146 |
→ Find user
→ Recent actions
Codeforces (c) Copyright 2010-2025 Mike Mirzayanov
The only programming contests Web 2.0 platform
Server time: Feb/03/2025 01:55:01 (k2).
Desktop version, switch to mobile version.
Supported by
Auto comment: topic has been updated by noobsolver24 (previous revision, new revision, compare).
Interesting, What makes you say this? I just tried it for problem C and it could not solve.
you can just try to solve d then it help you :)
i jus feel like similar problem appeared in past somewhere. thats why many ppl solved it and it could be a possibility that o3 had the problem data so it solved it too or iam pretty much coping ;/
fr
RIP the upcoming Div4.
it shouldn't be rated imo
This guy didnt even try to hide it :skull:
well the questions were already fucked. imo they don't match their elo
i think this is a diff ai model not o-3 mini , its written in diff style
M-m-m-massive????
I wander if it is really a div2 contest?
Check how the codestyle of the 6th place harsh__h is different in problem D and in other problems. That's insane if he cheated since he is CM/Master and apparently works at Google, and also was a problemsetter.
The different use of spaces (D has, and ABC don't), also '\n' in D instead of endl in ABC.
No, he didn't. I know him personally. He is my senior cum mentor. He has been giving contests from many years. D has a very well-known solving technique of using 2D dijikstra. Even I was able to think of the right solution within a couple of minutes. He perhaps had a similar template by which he was able to submit the solution within 3 mins. You just have to change the innermost loop inn dijikstra .i.e if dist[node1][node2] > prev_dist+abs(node1-node2). I didn't have any such template so it took me some time.
Nobody argues that D is straightforward. How do you explain the different codestyle?
Also, what do you mean templates? He has different codestyle even in input and output sections across the problems in this round, there are no templates in that.
By the way, you don't have to defend him. It is he who cheated not you.
The input is the same. Even I use while(t--) thing it's quite common. On basis of output which is just endl and \n you can't blame a reputed coder. He is a master and he could have just interchanged by mistake for '\n' and endl
You are either vision-impaired or just trying your ass off to defend your bro.
Not only '\n' and endl are different. He also either use NO spaces before each bracket in some problems or uses spaces before ALL brackets. Is this also just a mistake? Like, about hundreds places in code where he "mistakenly" didn't put space or vice versa?
So arguing that someone has put spaces means he has cheated. When u first put a space on a bracket then u visually set yourself to put space in every bracket thereafter. In the same way when u don't put space on the bracket initially then u thereafter don't put any spaces for the remaining bracket.This is normal human tendency
Moreover, these coders type very fast. They are not going to look after spaces.
I'm a bit angry about the matter of things with you guys.
I will try my best to write a detailed blog about the situation with harsh__h in the upcoming days. I hope I will find time for that.
U are unnecessarily wasting ur time to undermine somenone's reputation. Cmon man, u can practice codeforces and one day u can be like other masters instead of wasting ur time.
It is very necessary because he once was a problemsetter, and without public attention, he might become one in the future too. And this is outright wrong when a cheater is given access to problemsetting.
He is not a cheater. He didn't cheat.How come he solved E1 if he had cheated?Common sense dude. Why can't u accept that he did D after B. Then he realized he had a mistake on B. So he rectified it then submitted both. Now he is not going to delete the solution for D that he has already written.
stop defending the cheater bruh
He didn't cheat. How tf are u even blaming him on the based on spaces and difference of output statement? Even Mike is going to laugh at u guys.
And I would definitely prefer to solve some more problems instead of writing comments, which don't give anything to me personally. But for the sake of the community I'm willing to put some effort in making problemsetting more fair.
Ok fine for these kind of graph problems u can refer this.
https://codeforces.me/blog/entry/45897
This is the blog where I learned to solve 2d dijikstra.
There are plenty of questions here.
Not to mention the elephant in the room: he solved the problem in 3 min after problem B. This is just enough time to read and understand what the problem is asking. Notice it took him twice as long to solve A, which is even more straightforward. I think this is a strong evidence for cheating. That said, I bet nothing is going to happen.
No he submitted B within 15 mins then went to D as it was a graph problm without realizing he had a wrong submission on D.Then he rectified B and submitted D
senior cum mentor ? wtf is that?
LOL ASJDAKSJHDKJASDHKJASHDASDA
probably helps him at lonely nights
senior and mentor.Just google it man cum is used in place of and or '/'
It doesnt matter how standard the question is he aint reading all that and typing out that code in 3 minutes. Its chatgpt in action.
He directly switched from B to D after submitting B in 15 mins. He didn't realize he made a mistake in B and then he had already written up the code for D when he realized his mistake in B. Rectified it.Submitted B then D
bull shit. How tf do you directly go to D without even checking if B is accepted or not. lol.
Footage of a legendary grandmaster getting caught solving almost an entire problemset without even checking if A is accepted or not. lol.
Ok, so he confidently submitted B, then went for D and mid implementation he was like: wait, did I even get B? Came back, solved B correctly and then solved D? If you assume it took him 1 min fix B it is still 9 min to understand problem D and write the solution. It is possible, but notice he would also would have to jump problem C (maybe he just miraculously guessed D would be easier than C? Did he expend 1/2 min reading and understanding C before choosing to go to D, reducing time on problem D to 7 min?). I agree this problem is straightforward for anyone high ranked (the grandmasters and above solved in 9 min on average), but this is extremely suspicious considering he is not Nutella and also didn't carry this same performance on problem C.
I am not trying to say anything about the man's cheating case, I am just saying that skipping to the next problem without checking the verdict happens very often. And I don't think this has exceptions based on rating (unless they are cyan or less, or whatever).
Fair, this indeed happens
No he completed D then he saw he submitted wrong solution of B. Rectified it. Submitted B. Waited 2 mins for B to get finally accepted. Then went on to submit D on 3rd minute.
Makes sense; I checked his submissions of previous questions and I don't think there is conclusive evidence that he cheated to be completely honest. I won't be risking accusing a righteous man without solid evidence so: really sorry for accusing you of cheating harsh__h. Congratulations on your performance today!
Please see my historical performances too, you will find numerous under rank 50 performances, many with weird solving order. I have given almost infinite number of contests sometimes I get lucky, get good rank then fall back. This time i could have got 2nd rank if i would have fixed my E2 just 1 min before. The code for D isn't even big its just input taking and two loops.
Makes sense bro, sorry. Also, I guess a lot of people are upset about the actual cheaters (you can see some grey solving A and D only), so it kinda became witch hunting.
Beacuse he was confident even I do that for problem A and B.He is masters rated He wont expect getting a wrong solution on B so he skipped to D.
"he won't expect getting a wrong solution on B"
yet he had a wrong submission on A..
Yeah but how often do people of master caliber have wrong solutions on A and B. So you can't be so specific. Even I was halfway onto C when I realized my B was wrong. And moreover, the solutions get queued for 2-3 mins. U can't wait to look at the queue wasting your time
"master caliber" you really are brainwashed
Just see the number of problems he has solved. He had been master last year and there was not a huge buzz about AI or Chatgpt during that time.So u can say he did it all by himself.
And now about brainwashing stuff ,I have given few group contests (NOT CF) with him.I was myself a 1700 rated guy these days. He solved 4-5 problems out of which I did only 1.U are the one believing hard work cant achieve master's rating.
Now if u say why didn't do C after B, thats beacuse these top coders are quite good and confident on tress and graph because they have standard problems.
If u want I can give you more questions and links for this kind of problem using 2D or higher order dijikstra.
Please compare D with E1, E2 code too.
E1 E2 submitted during contest E2 fixed with a minor change.
I sometimes use autofill sometimes formatter sometimes i don't do anything. And didn't even take less time in D see my B time, reading that hell of a D problem statement itself took 3 min. I don't know how cf judges AI solutions but please run that evals on my solution and make sure I did not cheat.
Agree. The worst rount of CF. And also for me. My contest was very bad. I have been feeling very bad for two days. :([
have you tried all contests in CF that you say this one is "worst"?
u are right bro this round is little bit tough and its require a good understanding and practice to solve