Hi fellow kagglers,
\nI have noticed some interesting pattern in kernels for this competition - the majority of them have much more forks than upvotes. Here are some examples:
\nAs usual in any competition, I would like to ask you to remember the important point - if you like some kernel or fork it to create your own solution, it will be great to put upvote for the initial kernel and make the reference to it inside the new one. Upvoting is the great chance to say \"I respect your work\".
\nI hope this will help us to make the community relationship and experience exchange atmosphere even better
", "messages": [ { "id": "2492502", "postDate": "10/22/2023 15:24:05", "content": "Hi fellow kagglers,
\nI have noticed some interesting pattern in kernels for this competition - the majority of them have much more forks than upvotes. Here are some examples:
\nAs usual in any competition, I would like to ask you to remember the important point - if you like some kernel or fork it to create your own solution, it will be great to put upvote for the initial kernel and make the reference to it inside the new one. Upvoting is the great chance to say \"I respect your work\".
\nI hope this will help us to make the community relationship and experience exchange atmosphere even better
", "rawMarkdown": "Hi fellow kagglers,\n\nI have noticed some interesting pattern in kernels for this competition - the majority of them have much more forks than upvotes. Here are some examples:\n- [Zzzs: Random Forest model starter](https://www.kaggle.com/code/carlmcbrideellis/zzzs-random-forest-model-starter) by [@carlmcbrideellis](https://www.kaggle.com/carlmcbrideellis) with 158 upvotes and 672 forks (**23.5%** upvotes)\n- [Feature Engineering and Random Forest Prediction](https://www.kaggle.com/code/lccburk/feature-engineering-and-random-forest-prediction) by [@lccburk](https://www.kaggle.com/lccburk) with 117 upvotes and 528 forks (**22.16%** upvotes)\n- [Sleep - Critical point - Infer](https://www.kaggle.com/code/werus23/sleep-critical-point-infer) by [@werus23](https://www.kaggle.com/werus23) with only 85 upvotes and 492 forks (**17.2%** upvotes - the \"luckiest\" person)\n- [Detect Sleep States | Baseline Solution](https://www.kaggle.com/code/zhukovoleksiy/detect-sleep-states-baseline-solution) by [@zhukovoleksiy](https://www.kaggle.com/zhukovoleksiy) with 78 upvotes and 417 forks (**18.7%** upvotes)\n- [Explained Baseline Solution đź’¨](https://www.kaggle.com/code/zulqarnainali/explained-baseline-solution) by [@zulqarnainali](https://www.kaggle.com/zulqarnainali) with 57 upvotes and 146 forks (**39.04%** upvotes)\n\nAs usual in any competition, I would like to ask you to remember the important point - **if you like some kernel or fork it to create your own solution, it will be great to put upvote for the initial kernel and make the reference to it inside the new one. Upvoting is the great chance to say \"I respect your work\"**.\n\nI hope this will help us to make the community relationship and experience exchange atmosphere even better", "votes": null }, { "id": "2492520", "postDate": "10/22/2023 15:37:40", "content": "Exactly. The problem appears in every competition and there are notebooks that do not exceed 10% upvotes to fork. \n\"Before you fork, upvote!\"
", "rawMarkdown": "Exactly. The problem appears in every competition and there are notebooks that do not exceed 10% upvotes to fork. \n\"Before you fork, upvote!\"", "votes": null }, { "id": "2492528", "postDate": "10/22/2023 15:42:46", "content": "Yep @michau96, and this is our challenge to make the things better by using the method \"say it loud if you want to change something\"
", "rawMarkdown": "Yep @michau96, and this is our challenge to make the things better by using the method \"say it loud if you want to change something\"", "votes": null }, { "id": "2492632", "postDate": "10/22/2023 17:36:59", "content": "I agree. These are some really helpful notebooks so I don't understand why someone would use them and not upvote them prior to usage. I wish one could upvote more than once. Learnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL. That's why I love these competitions. You can't get that level of exposure even on some work projects.
", "rawMarkdown": "I agree. These are some really helpful notebooks so I don't understand why someone would use them and not upvote them prior to usage. I wish one could upvote more than once. Learnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL. That's why I love these competitions. You can't get that level of exposure even on some work projects.", "votes": null }, { "id": "2492642", "postDate": "10/22/2023 17:46:51", "content": "Yep @aidlre001, that's why I'm still on Kaggle after 12 years and now I have changed my position from \"receiving experience\" to \"exchange and share experience\". As not only me doing this, Kaggle is still alive and pretty active platform with a lot of great DS professionals
", "rawMarkdown": "Yep @aidlre001, that's why I'm still on Kaggle after 12 years and now I have changed my position from \"receiving experience\" to \"exchange and share experience\". As not only me doing this, Kaggle is still alive and pretty active platform with a lot of great DS professionals", "votes": null }, { "id": "2492657", "postDate": "10/22/2023 17:55:49", "content": "@Alexander Ryzhkov,
\nMy only regret is that I didn't join sooner. Though no choice due to work and lots of offline commitments and constraints. But I forced myself no matter what. Its a great platform. I just wish I had more access to larger GPUs and TPUs with more cores, RAM, etc, without the financial cost implications among others. Maybe some billionaire would come and combine Kaggle frameworks with a future machine/brain interface with the Metaverse and we could perform all our work without needing to log off and I would be running notebooks while I sleep even with power outages or bad network connections. Maybe that's a very bad idea but still…
", "rawMarkdown": "Alexander Ryzhkov,\n\nMy only regret is that I didn't join sooner. Though no choice due to work and lots of offline commitments and constraints. But I forced myself no matter what. Its a great platform. I just wish I had more access to larger GPUs and TPUs with more cores, RAM, etc, without the financial cost implications among others. Maybe some billionaire would come and combine Kaggle frameworks with a future machine/brain interface with the Metaverse and we could perform all our work without needing to log off and I would be running notebooks while I sleep even with power outages or bad network connections. Maybe that's a very bad idea but still...", "votes": null }, { "id": "2492681", "postDate": "10/22/2023 18:32:46", "content": "Hi Ryzhvov,
\nThe problem is bigger than forking and Not voting. Newbies aren't providing the due appreciation to THE REAL professionals (GMs or Not).
\nAll Notebooks on that list are submissions. Unfortunately, users forked it (and didn't vote) to be on LB.
\nIt's sad, Kaggle's Audience is changing ( not for good!)
\nNew kagglers intend to achieve positions WITHOUT any background.
\nCheck that: LLM Science Exam Competition. First Place Solution
\nJust a Bronze for their Notebook.
https://www.kaggle.com/code/ybabakhin/1st-place-team-h2o-llm-studio
\nOr maybe a Silver: https://www.kaggle.com/code/ybabakhin/1st-place-single-model-inference
\nNewbies want to reach anything/visibility fast and easy. However, that path simply doesn't exist.
", "rawMarkdown": "Hi Ryzhvov,\n\nThe problem is bigger than forking and Not voting. Newbies aren't providing the due appreciation to THE REAL professionals (GMs or Not).\n\nAll Notebooks on that list are submissions. Unfortunately, users forked it (and didn't vote) to be on LB.\n\nIt's sad, Kaggle's Audience is changing ( not for good!)\n\nNew kagglers intend to achieve positions WITHOUT any background. \n\nCheck that: LLM Science Exam Competition. First Place Solution\nJust a Bronze for their Notebook.\n\nhttps://www.kaggle.com/code/ybabakhin/1st-place-team-h2o-llm-studio\n\nOr maybe a Silver: https://www.kaggle.com/code/ybabakhin/1st-place-single-model-inference\n\nNewbies want to reach anything/visibility fast and easy. However, that path simply doesn't exist.", "votes": null }, { "id": "2492839", "postDate": "10/23/2023 00:01:41", "content": "Indeed, looking at these numbers, it seems that not all people who have not voted in favor are new to Kaggle. This is a serious etiquette issue. Also here is @ravi20076's great solution, please feel free to \"upvote\".
\nhttps://www.kaggle.com/discussions/product-feedback/442921#2462509
\n\nLearnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL.
\n
Totally agree with you đź’Ż on this point @aidlre001
", "rawMarkdown": "> Learnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL.\n\nTotally agree with you đź’Ż on this point @aidlre001", "votes": null }, { "id": "2492880", "postDate": "10/23/2023 02:05:51", "content": "Should clicking on Copy + Edit automatically upvote the notebook? What could be pros/cons of this?
This is unfortunately true for most competitions. For example, in the current Writing Quality competition, my 0.604 notebook (https://www.kaggle.com/code/mcpenguin/writing-processes-to-quality-baseline) has 53 upvotes but 161 forks (which is around the same percentage as the sleep notebooks you posted above).
\nAdditionally to my dismay someone copied my notebook, changing only the model parameters, got a higher LB score (0.602), and just republished it without saying the notebook originally came from me! I left a comment on that notebook, but it's extremely discouraging to see this happen to my work.
", "rawMarkdown": "This is unfortunately true for most competitions. For example, in the current Writing Quality competition, my 0.604 notebook (https://www.kaggle.com/code/mcpenguin/writing-processes-to-quality-baseline) has 53 upvotes but 161 forks (which is around the same percentage as the sleep notebooks you posted above). \n\nAdditionally to my dismay someone copied my notebook, changing only the model parameters, got a higher LB score (0.602), and just republished it without saying the notebook originally came from me! I left a comment on that notebook, but it's extremely discouraging to see this happen to my work.", "votes": null }, { "id": "2493593", "postDate": "10/23/2023 13:45:56", "content": "@bkowshik The pros is better upvoting score but the cons is the visibility of people who forked in private. The better version of that is to use just the number upvotes+forks_count for medals and ranking
This may be hard to add but it would be nice to have a reference or citation feature for each notebook as well as upvotes so once someone forkmits a notebook with or without upvoting or acknowledging the original poster, it would show \"cited/referenced from . Then the authors can actually generate reports or pdf documents of their notebooks with the number of citations, upvotes, etc. It would be really helpful for both authors and users alike. Especially for people like me who spend days or weeks trying to understand why and how the various solutions came about. Still trying to understand the thought processes and I think that should be the main takeaway (at least for me) from these competitions. This is how (at least) technical education should be, not from outdated textbooks and code snippets online.
", "rawMarkdown": "This may be hard to add but it would be nice to have a reference or citation feature for each notebook as well as upvotes so once someone forkmits a notebook with or without upvoting or acknowledging the original poster, it would show \"cited/referenced from@Marcus Chan,
\nPlease don't be discouraged. A lot of folks appreciate works that you guys do. It takes a lot of commitment to sit down and break down complex or technical ideas or solutions for free and yes, in the human world, some (ok, most) would take advantage of that but they are losing in the end because they don't grow. Those that acknowledge and give feedback actually may inspire new ideas for the original authors in ways they didn't even anticipate and both learn new things. These experiences will count in the real world every time based on experience.
", "rawMarkdown": "Marcus Chan,\n\nPlease don't be discouraged. A lot of folks appreciate works that you guys do. It takes a lot of commitment to sit down and break down complex or technical ideas or solutions for free and yes, in the human world, some (ok, most) would take advantage of that but they are losing in the end because they don't grow. Those that acknowledge and give feedback actually may inspire new ideas for the original authors in ways they didn't even anticipate and both learn new things. These experiences will count in the real world every time based on experience.", "votes": null }, { "id": "2493794", "postDate": "10/23/2023 15:03:38", "content": "@mcpenguin here the quick idea could be like this - all the kernel forks have the H1 disclaimer in the beginning with the link to the parent kernel
", "rawMarkdown": "mcpenguin here the quick idea could be like this - all the kernel forks have the H1 disclaimer in the beginning with the link to the parent kernel", "votes": null }, { "id": "2493815", "postDate": "10/23/2023 15:15:27", "content": "@alexryzhkov Unfortunately this doesn’t work as is because forks of forks exist.
\nFor example, in thr writing quality competition, someone copied (and cited) my notebook and got a LB score of 0.603. From there, someone copied their notebook and got a LB score of 0.602, without citing my notebook.
\nInstead, we should mention not just the immediate parent, but also the “history” of forks (so back to me, and the notebooks I copied from as well). Just saying we should take this into account.
", "rawMarkdown": "alexryzhkov Unfortunately this doesn’t work as is because forks of forks exist.\n\nFor example, in thr writing quality competition, someone copied (and cited) my notebook and got a LB score of 0.603. From there, someone copied their notebook and got a LB score of 0.602, without citing my notebook.\n\nInstead, we should mention not just the immediate parent, but also the “history” of forks (so back to me, and the notebooks I copied from as well). Just saying we should take this into account.", "votes": null }, { "id": "2493819", "postDate": "10/23/2023 15:16:37", "content": "Would someone be able to give two “upvotes” by upvoting and forking? I feel it should maybe be that a user can only “upvote” once.
", "rawMarkdown": "Would someone be able to give two “upvotes” by upvoting and forking? I feel it should maybe be that a user can only “upvote” once.", "votes": null }, { "id": "2503346", "postDate": "10/29/2023 02:06:12", "content": "Is there any disadvantage to upvote? No, there isn't, is there?
\nI keep upvoting in favor of discussion and notebook that I find helpful.
\nI am more cautious about forking. It will only increase my code.
\nMore respect should be given to useful notebook and discussion.
@minfuka thanks for your opinion!
\nI also can’t see any problem in upvoting. As for forks - for me it is the option for better understanding the created code. You can run it with different params, check the outputs of the specific functions and so on. But you are right - it should not become a new public kernel without any disclaimer
", "rawMarkdown": "minfuka thanks for your opinion!\n\nI also can’t see any problem in upvoting. As for forks - for me it is the option for better understanding the created code. You can run it with different params, check the outputs of the specific functions and so on. But you are right - it should not become a new public kernel without any disclaimer", "votes": null } ], "comments": [ { "id": 2492520, "author_name": "michau96", "author_url": "", "post_date": "10/22/2023 15:37:40", "content": "Exactly. The problem appears in every competition and there are notebooks that do not exceed 10% upvotes to fork. \n\"Before you fork, upvote!\"
", "votes": null, "replies": [ { "id": 2492528, "author_name": "alexryzhkov", "author_url": "", "post_date": "10/22/2023 15:42:46", "content": "Yep @michau96, and this is our challenge to make the things better by using the method \"say it loud if you want to change something\"
", "votes": null, "replies": [] } ] }, { "id": 2492632, "author_name": "aidlre001", "author_url": "", "post_date": "10/22/2023 17:36:59", "content": "I agree. These are some really helpful notebooks so I don't understand why someone would use them and not upvote them prior to usage. I wish one could upvote more than once. Learnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL. That's why I love these competitions. You can't get that level of exposure even on some work projects.
", "votes": null, "replies": [ { "id": 2492642, "author_name": "alexryzhkov", "author_url": "", "post_date": "10/22/2023 17:46:51", "content": "Yep @aidlre001, that's why I'm still on Kaggle after 12 years and now I have changed my position from \"receiving experience\" to \"exchange and share experience\". As not only me doing this, Kaggle is still alive and pretty active platform with a lot of great DS professionals
", "votes": null, "replies": [ { "id": 2492657, "author_name": "aidlre001", "author_url": "", "post_date": "10/22/2023 17:55:49", "content": "@Alexander Ryzhkov,
\nMy only regret is that I didn't join sooner. Though no choice due to work and lots of offline commitments and constraints. But I forced myself no matter what. Its a great platform. I just wish I had more access to larger GPUs and TPUs with more cores, RAM, etc, without the financial cost implications among others. Maybe some billionaire would come and combine Kaggle frameworks with a future machine/brain interface with the Metaverse and we could perform all our work without needing to log off and I would be running notebooks while I sleep even with power outages or bad network connections. Maybe that's a very bad idea but still…
", "votes": null, "replies": [] } ] }, { "id": 2492878, "author_name": "bkowshik", "author_url": "", "post_date": "10/23/2023 01:57:36", "content": "\n\nLearnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL.
\n
Totally agree with you đź’Ż on this point @aidlre001
", "votes": null, "replies": [] } ] }, { "id": 2492681, "author_name": "mpwolke", "author_url": "", "post_date": "10/22/2023 18:32:46", "content": "Hi Ryzhvov,
\nThe problem is bigger than forking and Not voting. Newbies aren't providing the due appreciation to THE REAL professionals (GMs or Not).
\nAll Notebooks on that list are submissions. Unfortunately, users forked it (and didn't vote) to be on LB.
\nIt's sad, Kaggle's Audience is changing ( not for good!)
\nNew kagglers intend to achieve positions WITHOUT any background.
\nCheck that: LLM Science Exam Competition. First Place Solution
\nJust a Bronze for their Notebook.
https://www.kaggle.com/code/ybabakhin/1st-place-team-h2o-llm-studio
\nOr maybe a Silver: https://www.kaggle.com/code/ybabakhin/1st-place-single-model-inference
\nNewbies want to reach anything/visibility fast and easy. However, that path simply doesn't exist.
", "votes": null, "replies": [] }, { "id": 2492839, "author_name": "mitsuyasuhoshino", "author_url": "", "post_date": "10/23/2023 00:01:41", "content": "Indeed, looking at these numbers, it seems that not all people who have not voted in favor are new to Kaggle. This is a serious etiquette issue. Also here is @ravi20076's great solution, please feel free to \"upvote\".
\nhttps://www.kaggle.com/discussions/product-feedback/442921#2462509
Should clicking on Copy + Edit automatically upvote the notebook? What could be pros/cons of this?
@bkowshik The pros is better upvoting score but the cons is the visibility of people who forked in private. The better version of that is to use just the number upvotes+forks_count for medals and ranking
Would someone be able to give two “upvotes” by upvoting and forking? I feel it should maybe be that a user can only “upvote” once.
", "votes": null, "replies": [] } ] } ] }, { "id": 2493555, "author_name": "mcpenguin", "author_url": "", "post_date": "10/23/2023 13:29:21", "content": "This is unfortunately true for most competitions. For example, in the current Writing Quality competition, my 0.604 notebook (https://www.kaggle.com/code/mcpenguin/writing-processes-to-quality-baseline) has 53 upvotes but 161 forks (which is around the same percentage as the sleep notebooks you posted above).
\nAdditionally to my dismay someone copied my notebook, changing only the model parameters, got a higher LB score (0.602), and just republished it without saying the notebook originally came from me! I left a comment on that notebook, but it's extremely discouraging to see this happen to my work.
", "votes": null, "replies": [ { "id": 2493731, "author_name": "aidlre001", "author_url": "", "post_date": "10/23/2023 14:40:29", "content": "@Marcus Chan,
\nPlease don't be discouraged. A lot of folks appreciate works that you guys do. It takes a lot of commitment to sit down and break down complex or technical ideas or solutions for free and yes, in the human world, some (ok, most) would take advantage of that but they are losing in the end because they don't grow. Those that acknowledge and give feedback actually may inspire new ideas for the original authors in ways they didn't even anticipate and both learn new things. These experiences will count in the real world every time based on experience.
", "votes": null, "replies": [] }, { "id": 2493794, "author_name": "alexryzhkov", "author_url": "", "post_date": "10/23/2023 15:03:38", "content": "@mcpenguin here the quick idea could be like this - all the kernel forks have the H1 disclaimer in the beginning with the link to the parent kernel
", "votes": null, "replies": [ { "id": 2493815, "author_name": "mcpenguin", "author_url": "", "post_date": "10/23/2023 15:15:27", "content": "@alexryzhkov Unfortunately this doesn’t work as is because forks of forks exist.
\nFor example, in thr writing quality competition, someone copied (and cited) my notebook and got a LB score of 0.603. From there, someone copied their notebook and got a LB score of 0.602, without citing my notebook.
\nInstead, we should mention not just the immediate parent, but also the “history” of forks (so back to me, and the notebooks I copied from as well). Just saying we should take this into account.
", "votes": null, "replies": [] } ] } ] }, { "id": 2493722, "author_name": "aidlre001", "author_url": "", "post_date": "10/23/2023 14:35:26", "content": "This may be hard to add but it would be nice to have a reference or citation feature for each notebook as well as upvotes so once someone forkmits a notebook with or without upvoting or acknowledging the original poster, it would show \"cited/referenced from . Then the authors can actually generate reports or pdf documents of their notebooks with the number of citations, upvotes, etc. It would be really helpful for both authors and users alike. Especially for people like me who spend days or weeks trying to understand why and how the various solutions came about. Still trying to understand the thought processes and I think that should be the main takeaway (at least for me) from these competitions. This is how (at least) technical education should be, not from outdated textbooks and code snippets online.
", "votes": null, "replies": [] }, { "id": 2503346, "author_name": "minfuka", "author_url": "", "post_date": "10/29/2023 02:06:12", "content": "Is there any disadvantage to upvote? No, there isn't, is there?
\nI keep upvoting in favor of discussion and notebook that I find helpful.
\nI am more cautious about forking. It will only increase my code.
\nMore respect should be given to useful notebook and discussion.
@minfuka thanks for your opinion!
\nI also can’t see any problem in upvoting. As for forks - for me it is the option for better understanding the created code. You can run it with different params, check the outputs of the specific functions and so on. But you are right - it should not become a new public kernel without any disclaimer
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2492502": "Hi fellow kagglers,\n\nI have noticed some interesting pattern in kernels for this competition - the majority of them have much more forks than upvotes. Here are some examples:\n- [Zzzs: Random Forest model starter](https://www.kaggle.com/code/carlmcbrideellis/zzzs-random-forest-model-starter) by [@carlmcbrideellis](https://www.kaggle.com/carlmcbrideellis) with 158 upvotes and 672 forks (**23.5%** upvotes)\n- [Feature Engineering and Random Forest Prediction](https://www.kaggle.com/code/lccburk/feature-engineering-and-random-forest-prediction) by [@lccburk](https://www.kaggle.com/lccburk) with 117 upvotes and 528 forks (**22.16%** upvotes)\n- [Sleep - Critical point - Infer](https://www.kaggle.com/code/werus23/sleep-critical-point-infer) by [@werus23](https://www.kaggle.com/werus23) with only 85 upvotes and 492 forks (**17.2%** upvotes - the \"luckiest\" person)\n- [Detect Sleep States | Baseline Solution](https://www.kaggle.com/code/zhukovoleksiy/detect-sleep-states-baseline-solution) by [@zhukovoleksiy](https://www.kaggle.com/zhukovoleksiy) with 78 upvotes and 417 forks (**18.7%** upvotes)\n- [Explained Baseline Solution đź’¨](https://www.kaggle.com/code/zulqarnainali/explained-baseline-solution) by [@zulqarnainali](https://www.kaggle.com/zulqarnainali) with 57 upvotes and 146 forks (**39.04%** upvotes)\n\nAs usual in any competition, I would like to ask you to remember the important point - **if you like some kernel or fork it to create your own solution, it will be great to put upvote for the initial kernel and make the reference to it inside the new one. Upvoting is the great chance to say \"I respect your work\"**.\n\nI hope this will help us to make the community relationship and experience exchange atmosphere even better", "2492520": "Exactly. The problem appears in every competition and there are notebooks that do not exceed 10% upvotes to fork. \n\"Before you fork, upvote!\"", "2492528": "Yep @michau96, and this is our challenge to make the things better by using the method \"say it loud if you want to change something\"", "2492632": "I agree. These are some really helpful notebooks so I don't understand why someone would use them and not upvote them prior to usage. I wish one could upvote more than once. Learnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL. That's why I love these competitions. You can't get that level of exposure even on some work projects.", "2492642": "Yep @aidlre001, that's why I'm still on Kaggle after 12 years and now I have changed my position from \"receiving experience\" to \"exchange and share experience\". As not only me doing this, Kaggle is still alive and pretty active platform with a lot of great DS professionals", "2492657": "Alexander Ryzhkov,\n\nMy only regret is that I didn't join sooner. Though no choice due to work and lots of offline commitments and constraints. But I forced myself no matter what. Its a great platform. I just wish I had more access to larger GPUs and TPUs with more cores, RAM, etc, without the financial cost implications among others. Maybe some billionaire would come and combine Kaggle frameworks with a future machine/brain interface with the Metaverse and we could perform all our work without needing to log off and I would be running notebooks while I sleep even with power outages or bad network connections. Maybe that's a very bad idea but still...", "2492681": "Hi Ryzhvov,\n\nThe problem is bigger than forking and Not voting. Newbies aren't providing the due appreciation to THE REAL professionals (GMs or Not).\n\nAll Notebooks on that list are submissions. Unfortunately, users forked it (and didn't vote) to be on LB.\n\nIt's sad, Kaggle's Audience is changing ( not for good!)\n\nNew kagglers intend to achieve positions WITHOUT any background. \n\nCheck that: LLM Science Exam Competition. First Place Solution\nJust a Bronze for their Notebook.\n\nhttps://www.kaggle.com/code/ybabakhin/1st-place-team-h2o-llm-studio\n\nOr maybe a Silver: https://www.kaggle.com/code/ybabakhin/1st-place-single-model-inference\n\nNewbies want to reach anything/visibility fast and easy. However, that path simply doesn't exist.", "2492839": "Indeed, looking at these numbers, it seems that not all people who have not voted in favor are new to Kaggle. This is a serious etiquette issue. Also here is @ravi20076's great solution, please feel free to \"upvote\".\nhttps://www.kaggle.com/discussions/product-feedback/442921#2462509", "2492878": "> Learnt so many new more efficient libraries and techniques in this competition than from all the textbooks and online resources on ML and DL.\n\nTotally agree with you đź’Ż on this point @aidlre001", "2492880": "Should clicking on `Copy + Edit` automatically upvote the notebook? What could be pros/cons of this?", "2493555": "This is unfortunately true for most competitions. For example, in the current Writing Quality competition, my 0.604 notebook (https://www.kaggle.com/code/mcpenguin/writing-processes-to-quality-baseline) has 53 upvotes but 161 forks (which is around the same percentage as the sleep notebooks you posted above). \n\nAdditionally to my dismay someone copied my notebook, changing only the model parameters, got a higher LB score (0.602), and just republished it without saying the notebook originally came from me! I left a comment on that notebook, but it's extremely discouraging to see this happen to my work.", "2493593": "bkowshik The pros is better upvoting score but the cons is the visibility of people who forked in private. The better version of that is to use just the number `upvotes+forks_count` for medals and ranking", "2493722": "This may be hard to add but it would be nice to have a reference or citation feature for each notebook as well as upvotes so once someone forkmits a notebook with or without upvoting or acknowledging the original poster, it would show \"cited/referenced from