It’s worth noting that this athlete’s performance is quite impressive, with significant improvements over time.
While examining his code, I noticed some issues that can’t be fully proven, but which are still worth paying attention to.
This includes, but is not limited to, overly neat arrangements of variables, as well as whether there’s an excessive amount of WA1/CE/IDL1…
I’m curious whether this person is an official contestant who’s allowed to use AI in the competition, or if there’s any evidence that he isn’t using AI at all.
It’s just a possibility and a question that’s being raised.
Ummm I've looked at the submission history of this person. Quite definitely a cheater. And the entire existence of this account is just some loser trying to piss people off.
I would discuss any findings with an admin directly. If it seemed any would care to listen.
edit: I see in your most recent contest that this guy was directly above you in the rankings. condolences. I wish I could help report some of this garbage myself but honestly idk how. obviously it shouldn't be in public
XD, thks
I’m just wondering whether this is allowed or not.
why would it be allowed at all?
even if openai themselves wanted to test their LLM, it would be very bad to do it on a running contest and affect peoples' rating.
probably just some stup1d guy trying to unbalance the contests, where are the admins?
also congrats on GM ur lwk cracked
Why is it obvious this shouldn't be done in public?
If you can't figure that out yourself, there's no point in me explaining it to you.
Try figuring out the reason yourself :) If you can't, try asking your AI for help on it.
Not sure where is this passive aggressive hostility coming from, but whatever, I'll engage nevertheless.
It's really just not obvious to me why we shouldn't publicly berate these people. I've literally browsed for 15 minutes, and found around ~30 cheaters from random contests going months back, all Expert+. Now consider, I excluded the ones with subtle signs. Literally 1 out of every 3 account I've checked had gpt slapped on top of it. (I've only checked suspiciously meteoric ones but still)
So I ask once again, explain to me why is it obvious? Because they'll learn what to avoid? You can't catch smart ones anyway, and the stupid ones are too stupid to hide themselves. Not doing enough only powers more of these idiots.
The fact is, this is an incredibly frustrating platform, that makes you feel incredibly stupid at times. Seeing these just add insult to injury.
Also, it's a bit irrelevant but I feel the need to add this. This is a virus. Like I said, this is frustrating at times and plays on your ego, and eventually one can think "no way 1k ppl sovled this bro, i'll just gpt for hitns" and it spirals out of control like that. It kills entire games. Just read about what happened at cycling scene.
Public shaming does nothing. Do you not read the CF blogs ever? It happens constantly, and many of those accused people are continuing to get away with it.
The reason is that admin would have to read tons of garbage if they actually went through every time they were pinged on the blogs. Unfortunately, even many of the posts that call out cheaters are useless and don't provide real evidence.
Furthermore, a big reason why it's probably not productive to discuss in public why certain features of the code are probably signs of cheating, is that cheaters will then read it and know to avoid those signs! Already it has happened multiple times.
Forget the fact that many people that are willing to do research probably don't want their name publicly attached to accusations on the blog.
In general, public discussion has no positives and does have some negatives. I understand that you're upset about cheaters, me too.
What needs to happen is that people like you and me can find evidence to expose people ourselves, and be able securely to forward it to admin in a place where they'll see it. The community is powerful, and administrators need to learn how to use that.
lwk ban or at least mark in cheater database so other ppl know
also when is rollback and how tf does an AI hit top 20 in a Div 1 codeforces might be cooked ICPC and offline contests are the only true knowledge tests at this point.
I just reported this user in the database. Thanks
Also, when you look at the standings of recent div $$$1$$$ s, it seems like a lot of decent users (like $$$cm-m$$$) have also started cheating. Like sure, maybe some of them got to $$$gm$$$ or $$$igm$$$ in a few contests, but statistically, a good portion of them are cheating. I thought that like, at that level, you'd have enough respect for the game (and yourself) to not cheat, but I guess not. I wonder how many people in a given div $$$1$$$ are cheating in some way. I'd assume it's over $$$30\%$$$.
Anyway, to me it's kind of crazy that technology enabled this super cool thing called online participation for a lot of mind sports, but it only lasted like $$$20$$$ years before technology came back and ruined it. Maybe that will be the case with a lot of other online things, given that $$$AI$$$ is so good at breaking stuff now.
I feel like the only real way to continue having rating be a relevant tool to ones' ability to solve problems is to have codeforces run in a lockdown browser but then people will complain its too restrictive but unfortunately if people want to play stupid games we win stupid prizes
Such a solution would likely not be well received for the exact reason that you mentioned, and of course, because people can easily find other ways to circumvent the guardrails here (phone, for example).
Though, there's another point in your favour here — if anyone remembers it, as bad as TopCoder's applet felt, I'm sure only truly motivated people would've gone through the pain of setting it up, so, yeah, if the setup is painful, less cheaters, I guess?
Agree with you on the "respect for the game", sadly that isn't the case it seems. I haven't been active of late, but from what I hear, the situation's bonkers. I find it astonishing that people's integrity has such little cost that they're happy selling it for online points (ratings), but unfortunately, that's the case, and it affects honest participants.
what about bismispis? is he likely a cheater too?
Is he a CF duel user?Watch his name.
I hope the future of competitive programming isnot going to be busted because of the upgrades in AI and people using it like that.