Блог пользователя Pranav16

Автор Pranav16, история, 5 месяцев назад, По-английски

Hey Codeforces Community,

There's a user, chayankumawat2004, who seems to be cheating his way to Candidate Master rank. He uses gibberish language in his codes to prevent it from plagiarism, for example he made a debugger function, which is nothing but a bfs, with no node pushed, he also comments out templates, and sometimes, add gibberish code in main function even after printing the answer which is obviously cheating, and he is doing this to avoid plagiarism. When i confronted him, and asked for explanations, he couldn't answer and just tried to divert me from this topic.

He has been performing really well in recent contest, even today, he solved A-D, with the help of cheating, and gave almost master level performance

Here are some examples of his questionable submissions:

258916350

262776253

264959663

Despite his unethical behavior, he's still progressing towards CM and achieving good ranks. I am not jealous of something, but I am more of disheartened, that he is always able to make his way out easily.

I request moderators and MikeMirzayanov to take action against him

EDIT : As this blog has pointed out that Global Round 26 B was (perhaps intentionally) leaked and later failed on system testing due to a reason which makes cheating quite obvious to notice. This guy's submission also failed on the same test for the exact same reason.

  • Проголосовать: нравится
  • +177
  • Проголосовать: не нравится

»
5 месяцев назад, # |
  Проголосовать: нравится +42 Проголосовать: не нравится

he missed the happiness of getting AC from self hard work.

»
5 месяцев назад, # |
  Проголосовать: нравится +30 Проголосовать: не нравится

I think one should always leak a solution which fails on main tests. Better for the contest

  • »
    »
    5 месяцев назад, # ^ |
      Проголосовать: нравится 0 Проголосовать: не нравится

    that would require for the person to know the systests, perhaps an author?

  • »
    »
    5 месяцев назад, # ^ |
      Проголосовать: нравится +5 Проголосовать: не нравится

    I don't think it is possible to write a solution with completely wrong logic which pass all the pretests. Best you can do is add weird edge cases but that can be easily bypassed by understanding the logic of the wrong solution and then correcting it yourself

»
5 месяцев назад, # |
Rev. 2   Проголосовать: нравится +2 Проголосовать: не нравится

My friend sahild2003 likes to use ChatGPT when participating in contests. You can see it by his submission history in 10 different programming languages — https://codeforces.net/submissions/sahild2003

He does the same on Codechef and Leetcode as well, but these platforms haven't banned him yet. Pls at least ban him on Codeforces.

  • »
    »
    5 месяцев назад, # ^ |
    Rev. 2   Проголосовать: нравится +21 Проголосовать: не нравится

    I may be wrong , but as of now , i dont think using chat gpt is considered cheating unless stated anywhere.I am completely against cheating , but like atcoder abc introduces a rule to not copy paste it on prompt , i dont think any such rule has been given for codeforces, codechef or leetcode

    • »
      »
      »
      5 месяцев назад, # ^ |
      Rev. 2   Проголосовать: нравится 0 Проголосовать: не нравится

      Hmmm, Leetcode does't specify that ChatGPT is not allowed, but according to the rules, the following is prohibited: "Multiple accounts submitting similar code for the same problem.".

      Codeforces doesn't prohibit using ChatGPT explicitly, but it will detect a similar solutions for the problem. This can be easily avoided by changing variable names and restructuring the code I think.

      Should't the rules be adjusted to prohibit submitting solutions that are fully written by ChatGPT?

      • »
        »
        »
        »
        5 месяцев назад, # ^ |
          Проголосовать: нравится +11 Проголосовать: не нравится

        i dont think so

        Chatgpt can solve maybe 1 — 2 problem in Div2 at most (and often not even A), While a good coder can use it to his advantage (and imo not unfairly).

        For example, i dont have much experience with working with python (needed it for big int), so i solved AGC 066B by telling Chatgpt what i needed it to do

        Should we ban github copilot too?

        • »
          »
          »
          »
          »
          5 месяцев назад, # ^ |
            Проголосовать: нравится +4 Проголосовать: не нравится

          Even though ChatGPT obviously can't solve non-trivial problems by itself, it can be used to generate ideas.

          For instance, in this problem I forgot about the existence of Eulerian path and when I had a feeling that I was missing something I asked ChatGPT. After a few prompts it spit out a sentence with "Eulerian path" in it and it was enough for me

          In this problem I also didn't see something obvious and asked ChatGPT, tho I can't remember what it was

          Generally, ChatGPT can be used as a dumb friend who screams random words and eventually says something smart. I would consider this as an unfair usage

          • »
            »
            »
            »
            »
            »
            5 месяцев назад, # ^ |
              Проголосовать: нравится +11 Проголосовать: не нравится

            I think you may want to read Regarding Rule Changes in ABC Due to the Rise of Generative AI since you mentioned an AtCoder problem:

            Therefore, even if there are situations where using generative AI is advantageous in ARC/AGC, we judge that it does not reach a level that undermines competitiveness. From another perspective, we also believe that competing in problem-solving abilities, including the use of generative AI, has meaning.

            For example, when debugging your program, it is common to write a slow exhaustive search solution and compare outputs. In this case, if you use generative AI to write the exhaustive search, it is just a smart use of AI without depending on it to solve the problem.

            I share their sentiment.

        • »
          »
          »
          »
          »
          5 месяцев назад, # ^ |
          Rev. 2   Проголосовать: нравится 0 Проголосовать: не нравится

          Yeah, generating ideas for harder problems seems like a valid usage of ChatGPT and it might not yet be able to solve harder problems on CF. I think most of the problems are too verbose and ChatGPT gets confused.

          For Leetcode however, the problems are less verbose and similar to existing problems, so ChatGPT is able to solve at least 3 problems from the weekly contest and sometimes 4.

          Should we ban github copilot too?

          Copilot is based on GPT, so it is kinda the same thing

      • »
        »
        »
        »
        5 месяцев назад, # ^ |
          Проголосовать: нравится 0 Проголосовать: не нравится

        Indeed, I remember having a similar discussion in Is ChatGPT Actually Cheating? last year. Eventually, AtCoder published an update to their rules similar to the notion of cheating I developed from that LeetCode discussion. I think this solution is good enough for now.

    • »
      »
      »
      5 месяцев назад, # ^ |
        Проголосовать: нравится 0 Проголосовать: не нравится

      Hey Bro,Here's my opinion Even if it's not stated not to use AI Tools,see using them is a completely self loss Instead of practicing problems here we could just try to improve the AI models which can eventually solve more problems. Btw you have become pretty famous.

    • »
      »
      »
      5 месяцев назад, # ^ |
      Rev. 2   Проголосовать: нравится 0 Проголосовать: не нравится

      In this case its against the rules because it wasnt generated before the contest,https://codeforces.net/blog/entry/8790

      • »
        »
        »
        »
        5 месяцев назад, # ^ |
        Rev. 2   Проголосовать: нравится -8 Проголосовать: не нравится

        Yeah, these 2 points are a bit ambiguous:

        1. the code was written and published/distributed before the start of the round,

        2. the code is generated using tools that were written and published/distributed before the start of the round.

        ChatGPT model was trained before the round, but the code was generated during the round.

  • »
    »
    5 месяцев назад, # ^ |
      Проголосовать: нравится +1 Проголосовать: не нравится

    You are telling me ChatGPT can solve till E in a DIV 3, that too using PHP ;<

»
5 месяцев назад, # |
  Проголосовать: нравится +7 Проголосовать: не нравится

Cheating has been a common scenario these days. People are live streaming on YouTube and giving solutions while the contest is running.

»
5 месяцев назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

I am guessing he used those function to get a TLE three times on D of Global Round 26. Because the code in each submission looks almost identical and it went from 2000ms+ -> 77ms. However if his trick is to write gibberish so someone is not willing to look further into his code, then it is working.

»
5 месяцев назад, # |
  Проголосовать: нравится +39 Проголосовать: не нравится

Even kn_joshua uses these tactics, but he is not so good at it and sometimes gets caught, coincidently from the same college too. Although his codes are very fun to read, he explains his codes through comments to avoid plag, and it sometimes works

  • »
    »
    5 месяцев назад, # ^ |
    Rev. 2   Проголосовать: нравится 0 Проголосовать: не нравится

    lmao,i still dont think josh cheats but yeah, when i asked him about this he mentioned that he sent his code to his friend twice and ergo the skipped contests .i feel like mentioning that a part of chayans template belongs to my code. i have never had any communications with him regarding my template,I have never personally shared this with him ~~~~~ struct Node { long long data; Node* left; Node* right;

    Node(long long value) : data(value), left(nullptr), right(nullptr) {}

    };

    void debug2(Node* root) { if (!root) return;

    queue<Node*> q;
    q.push(root);
    
    while (!q.empty()) {
        Node* current = q.front();
        q.pop();
    
    
        if (current->left) q.push(current->left);
        if (current->right) q.push(current->right);
    }

    }

    void debug(long long iterations) { for (long long GAY = 0; GAY < iterations; ++GAY) { GAY = GAY++; GAY = GAY--; } }

    void deleteTree(Node* root) { if (!root) return;

    deleteTree(root->left);
    deleteTree(root->right);
    
    delete root;

    }

    void debuger(int x) { long long numNodes = 7; long long careConstant = 14; Node* root = new Node(1); queue<Node*> q; q.push(root);

    long long nodesCreated = 1;
    
    while (!q.empty() && nodesCreated < numNodes) {
        Node* current = q.front();
        q.pop();
    
        current->left = new Node(++nodesCreated);
        q.push(current->left);
    
        if (nodesCreated < numNodes) {
            current->right = new Node(++nodesCreated);
            q.push(current->right);
        }
    }
    debug2(root);
    debug(careConstant);
    
    deleteTree(root);

    ~~~~~ is a part of an antiplag code i use, made it using chat GPT, preventive measure against false positives and also because me and my peers give contests in the same room, so just in case to protect myself.241957045 first appearance of the code.

»
5 месяцев назад, # |
  Проголосовать: нравится +7 Проголосовать: не нравится

he should be ban MikeMirzayanov

»
5 месяцев назад, # |
  Проголосовать: нравится +1 Проголосовать: не нравится

how is it possible to get 126 rank (in his recent cf contest), by cheating can anybody explain? getting rank over 2k is justifiable but 126..

»
5 месяцев назад, # |
Rev. 2   Проголосовать: нравится +3 Проголосовать: не нравится

Just don't care whatever others do......if we are honest and sincere it will definitely pay one day

»
5 месяцев назад, # |
  Проголосовать: нравится -25 Проголосовать: не нравится

I doubt he is cheating, just bad code practices. Theres no way cheaters can solve D and leak the solutions. He even solved 5 question in one of the div 2. I just looked up his linkedin. It says IIT Kanpur CSE, i doubt he would have to cheat, given his credentials.

  • »
    »
    5 месяцев назад, # ^ |
    Rev. 2   Проголосовать: нравится +6 Проголосовать: не нравится

    You might be in a wrong impression. I am here at IIT Kanpur and I personally know him. This guy cares only about cheating. To expose him along with what has been mentioned in the blog I would like to add more. Look at his old ID kum.chayan.

    He cheated earlier as well. The only thing that back then his solutions got skipped because he was not aware of how to avoid getting caught. To do clean up he made new ID after knowing how to escape computer checks. He might have avoided the checks of plagiarism but by directly looking at his submissions with current ID it is clear that he cheats.

    To add I have myself seen him directly copying from telegram channels. This guy has no ethics at all he has prepared a resume with fake achievements including ranks which he never secured.

    • »
      »
      »
      5 месяцев назад, # ^ |
        Проголосовать: нравится 0 Проголосовать: не нравится

      Yeah checked it, damn I never knew guys with that kind of credentials resort to cheating. I held IITK in high regard due to its coding culture. Seems like most of them are farce, with only a genuine number of handfuls. Oh lord. how sophisticated the cheaters have got. Making their way to candidate master through cheating, SHAME.

      • »
        »
        »
        »
        5 месяцев назад, # ^ |
        Rev. 2   Проголосовать: нравится +10 Проголосовать: не нравится

        Actually what happens is that few such foolish people spoil the image of the entire college which consists of maximum of good people

    • »
      »
      »
      5 месяцев назад, # ^ |
        Проголосовать: нравится 0 Проголосовать: не нравится

      IITians cheat too?!! They are the most hardworking students I've heard of and never in the world did I think I would come across this.

      • »
        »
        »
        »
        5 месяцев назад, # ^ |
          Проголосовать: нравится +8 Проголосовать: не нравится

        IITians is a very wide range of people, and being an IITian only means you are moderately good at a certain type of test, it does not say anything about your moral qualifications, nor does it say you'll be good enough at competitive programming to not have to resort to cheating.

»
5 месяцев назад, # |
Rev. 4   Проголосовать: нравится +1 Проголосовать: не нравится

MikeMirzayanov , we can clearly see that he ( chayankumawat2004 ) cheated in problem B ( beacuse he submitted that incorrect spreaded solution and got WA on test 9) and maybe in other problems also then why he got positive rating please look into this matter

»
5 месяцев назад, # |
  Проголосовать: нравится +8 Проголосовать: не нравится

Why isn't he banned yet? MikeMirzayanov

»
5 месяцев назад, # |
  Проголосовать: нравится +7 Проголосовать: не нравится

MikeMirzayanov CodeForces Community will like to see this account banned and unrated.

»
5 месяцев назад, # |
Rev. 14   Проголосовать: нравится -7 Проголосовать: не нравится

mike need to find solutions asap

»
5 месяцев назад, # |
Rev. 2   Проголосовать: нравится -7 Проголосовать: не нравится

Great contest keep going

»
5 месяцев назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

Check out DeanMMenezes submissions and timings ..when compilation error or wrong answer occurs ,switching my language(within 1 min) never helped me...(just a speculation)

»
5 месяцев назад, # |
  Проголосовать: нравится +1 Проголосовать: не нравится

The fact that if he's got banned, nothing will stop him from making a new account and continue his journey. In my opinion, it is better to upgrade the judge system in order to prevent more cheaters instead of focusing on an actual individual.