Congrats to all who are making submissions -- it has been impressive to watch the Kaggle community make progress and openly share strategies!
\nA quick clarification on the Early Sharing Prizes.
\nWe have updated the eligibility condition (3) to state:
\n\n\nSubmissions should only make use of information publicly available before the
\ntemporal_cutoffdates provided with test sequences.
(The original eligibility condition provided a single cutoff date.)
\nFor more tips on how to interpret temporal_cutoff, please check out the notes on the competition data page.
Congrats to all who are making submissions -- it has been impressive to watch the Kaggle community make progress and openly share strategies!
\nA quick clarification on the Early Sharing Prizes.
\nWe have updated the eligibility condition (3) to state:
\n\n\nSubmissions should only make use of information publicly available before the
\ntemporal_cutoffdates provided with test sequences.
(The original eligibility condition provided a single cutoff date.)
\nFor more tips on how to interpret temporal_cutoff, please check out the notes on the competition data page.
thanks for the update.
\ni think this message should be pinned (for late kagglers who may come into the competition later)
", "rawMarkdown": "thanks for the update.\n\ni think this message should be pinned (for late kagglers who may come into the competition later)", "votes": null }, { "id": "3148910", "postDate": "03/13/2025 16:19:32", "content": "Hi there - We considered that but new entrants would see the updated language as their first experience with the competition.
", "rawMarkdown": "Hi there - We considered that but new entrants would see the updated language as their first experience with the competition.", "votes": null }, { "id": "3149371", "postDate": "03/14/2025 05:43:59", "content": "Question on condition (1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible?
\nIf the answer is yes, then isn't it will discourage people to share their solution early and defeat the propose of Early Sharing Prize?
\nIt will be unfair also because participants from different time zone will have to stay awake to bet their luck trying submitting at the earliest possible just after the cutoff time? (Probably many had reach the above the threshold by the cutoff date judging from current trend)
", "rawMarkdown": "Question on condition (1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible? \n\nIf the answer is yes, then isn't it will discourage people to share their solution early and defeat the propose of Early Sharing Prize?\n \nIt will be unfair also because participants from different time zone will have to stay awake to bet their luck trying submitting at the earliest possible just after the cutoff time? (Probably many had reach the above the threshold by the cutoff date judging from current trend)", "votes": null }, { "id": "3149973", "postDate": "03/14/2025 22:43:01", "content": "\n\n(1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible?
\n
Disclaimer: I'm not the host. Ultimately the host set the rules. But from a game theoretic point of view, here're my best guesses.
\nIt's probably OK to publish beforehand. Here's why.
\nFrom the host's perspective, it's better that everyone shares every work/code from day 0 and nothing is held back. That way, people can build on each other's work/ideas and ultimately a better best solution will come about when the competition closes. Remember, the host doesn't care who wins. They only care how good (or bad) the final best solution is.
\nFrom a competitor's perspective, they don't care how good (or bad) the final solution is, as long as their solution is the best among all the competitors'. So, they have incentive not to share their work, to keep others' solutions as inferior to theirs as possible.
\nSo, I think the purpose of Early Sharing Prize is to encourage sharing of work/ideas. Therefore, to the host, if a competitor wants to publish earlier, so much the better.
\nThe only problem I see though, is what if teams A, B, and C all published their notebooks before the data refresh, and all of them end up scoring above human baseline?
", "rawMarkdown": ">(1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible?\n\nDisclaimer: I'm not the host. Ultimately the host set the rules. But from a game theoretic point of view, here're my best guesses.\n\nIt's probably OK to publish beforehand. Here's why.\n\nFrom the host's perspective, it's better that everyone shares every work/code from day 0 and nothing is held back. That way, people can build on each other's work/ideas and ultimately a better best solution will come about when the competition closes. Remember, the host doesn't care who wins. They only care how good (or bad) the final best solution is.\n\nFrom a competitor's perspective, they don't care how good (or bad) the final solution is, as long as *their* solution is the best among all the competitors'. So, they have incentive *not* to share their work, to keep others' solutions as inferior to theirs as possible.\n\nSo, I think the purpose of Early Sharing Prize is to encourage sharing of work/ideas. Therefore, to the host, if a competitor wants to publish earlier, so much the better.\n\nThe only problem I see though, is what if teams A, B, and C all published their notebooks before the data refresh, and all of them end up scoring above human baseline?", "votes": null } ], "comments": [ { "id": 3148906, "author_name": "hengck23", "author_url": "", "post_date": "03/13/2025 16:16:37", "content": "thanks for the update.
\ni think this message should be pinned (for late kagglers who may come into the competition later)
", "votes": null, "replies": [ { "id": 3148910, "author_name": "maggiemd", "author_url": "", "post_date": "03/13/2025 16:19:32", "content": "Hi there - We considered that but new entrants would see the updated language as their first experience with the competition.
", "votes": null, "replies": [] } ] }, { "id": 3149371, "author_name": "chewkokwahibrainai", "author_url": "", "post_date": "03/14/2025 05:43:59", "content": "Question on condition (1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible?
\nIf the answer is yes, then isn't it will discourage people to share their solution early and defeat the propose of Early Sharing Prize?
\nIt will be unfair also because participants from different time zone will have to stay awake to bet their luck trying submitting at the earliest possible just after the cutoff time? (Probably many had reach the above the threshold by the cutoff date judging from current trend)
", "votes": null, "replies": [ { "id": 3149973, "author_name": "revealer", "author_url": "", "post_date": "03/14/2025 22:43:01", "content": "\n\n(1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible?
\n
Disclaimer: I'm not the host. Ultimately the host set the rules. But from a game theoretic point of view, here're my best guesses.
\nIt's probably OK to publish beforehand. Here's why.
\nFrom the host's perspective, it's better that everyone shares every work/code from day 0 and nothing is held back. That way, people can build on each other's work/ideas and ultimately a better best solution will come about when the competition closes. Remember, the host doesn't care who wins. They only care how good (or bad) the final best solution is.
\nFrom a competitor's perspective, they don't care how good (or bad) the final solution is, as long as their solution is the best among all the competitors'. So, they have incentive not to share their work, to keep others' solutions as inferior to theirs as possible.
\nSo, I think the purpose of Early Sharing Prize is to encourage sharing of work/ideas. Therefore, to the host, if a competitor wants to publish earlier, so much the better.
\nThe only problem I see though, is what if teams A, B, and C all published their notebooks before the data refresh, and all of them end up scoring above human baseline?
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "3148857": "Congrats to all who are making submissions -- it has been impressive to watch the Kaggle community make progress and openly share strategies!\n\nA quick clarification on the [Early Sharing Prizes](https://www.kaggle.com/competitions/stanford-rna-3d-folding/overview). \n\nWe have updated the eligibility condition (3) to state:\n\n>Submissions should only make use of information publicly available before the `temporal_cutoff` dates provided with test sequences.\n\n(The original eligibility condition provided a single cutoff date.) \n\nFor more tips on how to interpret `temporal_cutoff`, please check out the notes on the [competition data](https://www.kaggle.com/competitions/stanford-rna-3d-folding/data) page.", "3148906": "thanks for the update.\n\ni think this message should be pinned (for late kagglers who may come into the competition later)", "3148910": "Hi there - We considered that but new entrants would see the updated language as their first experience with the competition.", "3149371": "Question on condition (1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible? \n\nIf the answer is yes, then isn't it will discourage people to share their solution early and defeat the propose of Early Sharing Prize?\n \nIt will be unfair also because participants from different time zone will have to stay awake to bet their luck trying submitting at the earliest possible just after the cutoff time? (Probably many had reach the above the threshold by the cutoff date judging from current trend)", "3149973": ">(1): What if someone publish an above benchmark notebook before the data refresh, is it consider illegible?\n\nDisclaimer: I'm not the host. Ultimately the host set the rules. But from a game theoretic point of view, here're my best guesses.\n\nIt's probably OK to publish beforehand. Here's why.\n\nFrom the host's perspective, it's better that everyone shares every work/code from day 0 and nothing is held back. That way, people can build on each other's work/ideas and ultimately a better best solution will come about when the competition closes. Remember, the host doesn't care who wins. They only care how good (or bad) the final best solution is.\n\nFrom a competitor's perspective, they don't care how good (or bad) the final solution is, as long as *their* solution is the best among all the competitors'. So, they have incentive *not* to share their work, to keep others' solutions as inferior to theirs as possible.\n\nSo, I think the purpose of Early Sharing Prize is to encourage sharing of work/ideas. Therefore, to the host, if a competitor wants to publish earlier, so much the better.\n\nThe only problem I see though, is what if teams A, B, and C all published their notebooks before the data refresh, and all of them end up scoring above human baseline?" }, "source": "meta" }