Hi,
\nI competed as a solo participant under the team name \"Santiago Maniches.\" My final private leaderboard position is 504 out of 1867, with a private score of 0.39968. My public leaderboard position is 404, with a public score of 0.42470.
\nI am attaching three screenshots that explain the situation better than words can.
\nThe first screenshot shows my submission history. Six submissions from within the deadline period each display \"Succeeded\" alongside a warning icon. All six have both public and private scores computed. Their private scores range from 0.502 to 0.544. The first place team on the private leaderboard finished at 0.49669, so any of these would have placed above the winner. The highest, at 0.54468, is above first place by approximately 0.048.
\nThe second and third screenshots show my private and public leaderboard positions, with the URL visible for identification. The public screenshot also shows Kaggle congratulating me on 0.42470 as my best entry. That score came from two notebook reruns Kaggle executed on March 31, which are the only submissions marked COMPLETE in my history. Those are what went into the final evaluation.
\nI am not writing to contest my placement or ask for anything to be reopened. The competition is closed and I understand the results are final. What I would like to understand is what the warning icon means on those six submissions. They were submitted before the deadline, Kaggle computed private scores for them, yet they were not included in the final evaluation.\n**\nThe reason this matters to me beyond the leaderboard is that this is real structural biology work. If those predictions were evaluated against the test set and represent a genuine quality difference, the methodology might be scientifically useful to the Stanford team independently of the competition outcome. The approach involved multi-template alignment with topological diversity selection across 5,716 competition templates and 167K Ribonanza structures, with constraint refinement passes for bond geometry and steric clashes.**
\nThree questions if someone can help:
\nKaggle Suport shared that \"Usually that refers to an error in the submission.\". However, my submissions are scored using both set the public and the hidden (private).
\nI am happy to share the relevant kernel and technical details privately.
\nThank you,
\nSantiago Maniches\nhttps://www.kaggle.com/santmani
", "messages": [ { "id": 3455654, "postDate": "2026-05-10T00:52:55.480Z", "content": "You selected two submissions. After the competition closed, all submissions are marked by the warning icon to indicate that they are pre-deadline submissions, and those private scores you see there are computed using a placeholder private set (which is different from the final private test set, so they mean nothing in terms of final placement). Then, the two selected submissions were rerun using the new test data. so you should see two green submissions on top of the submissions history (rerun) which means the results of them are used to determine your final standing (public score and private score you now see on LB). Hope that helps! I was confused in the beginning as well
", "rawMarkdown": "You selected two submissions. After the competition closed, all submissions are marked by the warning icon to indicate that they are pre-deadline submissions, and those private scores you see there are computed using a placeholder private set (which is different from the final private test set, so they mean nothing in terms of final placement). Then, the two selected submissions were rerun using the new test data. so you should see two green submissions on top of the submissions history (rerun) which means the results of them are used to determine your final standing (public score and private score you now see on LB). Hope that helps! I was confused in the beginning as well", "votes": 1, "replies": [ { "id": 3455666, "postDate": "2026-05-10T02:45:38.107Z", "content": "Thank you for your response. Got it ! I thought these were being scored against the entire public and private data sets, and if that was the case (considering the scoring), I wanted to share the notebook with the organizers. Thanks again for the clarification!
", "rawMarkdown": "Thank you for your response. Got it ! I thought these were being scored against the entire public and private data sets, and if that was the case (considering the scoring), I wanted to share the notebook with the organizers. Thanks again for the clarification!" } ] }, { "id": 3455040, "postDate": "2026-05-08T13:26:11.687Z", "content": "Hi @sohiermse , @ryanholbrook and @wcukierski
\nSince the competition closed a while ago, I am tagging some members of Kaggle team, to see if this situation is worth further investigation. Thanks in advance for your time and sorry for tagging you directly, I am hoping the message to be discovered. Thank you again and sorry for the inconvenience!
", "rawMarkdown": "Hi @sohiermse , @ryanholbrook and @wcukierski \n\nSince the competition closed a while ago, I am tagging some members of Kaggle team, to see if this situation is worth further investigation. Thanks in advance for your time and sorry for tagging you directly, I am hoping the message to be discovered. Thank you again and sorry for the inconvenience!", "replies": [ { "id": 3459945, "postDate": "2026-05-18T13:23:56.750Z", "content": "Sorry for the delay. The response by @honganzhu is correct. Generally, we add a note saying why submissions were invalidated (e.g., due to a rerun), but that step was missed, which added to the confusion. We've updated some process steps to make things more clear in the future.
", "rawMarkdown": "Sorry for the delay. The response by @honganzhu is correct. Generally, we add a note saying why submissions were invalidated (e.g., due to a rerun), but that step was missed, which added to the confusion. We've updated some process steps to make things more clear in the future.", "votes": 1, "replies": [ { "id": 3459950, "postDate": "2026-05-18T13:28:57.897Z", "content": "That's no problem. It makes sense now, thank you so much for the clarification.
", "rawMarkdown": "That's no problem. It makes sense now, thank you so much for the clarification. " } ] } ] }, { "id": 3455038, "postDate": "2026-05-08T13:11:23.457Z", "content": "Hi,
\nI competed as a solo participant under the team name \"Santiago Maniches.\" My final private leaderboard position is 504 out of 1867, with a private score of 0.39968. My public leaderboard position is 404, with a public score of 0.42470.
\nI am attaching three screenshots that explain the situation better than words can.
\nThe first screenshot shows my submission history. Six submissions from within the deadline period each display \"Succeeded\" alongside a warning icon. All six have both public and private scores computed. Their private scores range from 0.502 to 0.544. The first place team on the private leaderboard finished at 0.49669, so any of these would have placed above the winner. The highest, at 0.54468, is above first place by approximately 0.048.
\nThe second and third screenshots show my private and public leaderboard positions, with the URL visible for identification. The public screenshot also shows Kaggle congratulating me on 0.42470 as my best entry. That score came from two notebook reruns Kaggle executed on March 31, which are the only submissions marked COMPLETE in my history. Those are what went into the final evaluation.
\nI am not writing to contest my placement or ask for anything to be reopened. The competition is closed and I understand the results are final. What I would like to understand is what the warning icon means on those six submissions. They were submitted before the deadline, Kaggle computed private scores for them, yet they were not included in the final evaluation.\n**\nThe reason this matters to me beyond the leaderboard is that this is real structural biology work. If those predictions were evaluated against the test set and represent a genuine quality difference, the methodology might be scientifically useful to the Stanford team independently of the competition outcome. The approach involved multi-template alignment with topological diversity selection across 5,716 competition templates and 167K Ribonanza structures, with constraint refinement passes for bond geometry and steric clashes.**
\nThree questions if someone can help:
\nKaggle Suport shared that \"Usually that refers to an error in the submission.\". However, my submissions are scored using both set the public and the hidden (private).
\nI am happy to share the relevant kernel and technical details privately.
\nThank you,
\nSantiago Maniches\nhttps://www.kaggle.com/santmani
", "rawMarkdown": "Hi,\n\nI competed as a solo participant under the team name \"Santiago Maniches.\" My final private leaderboard position is 504 out of 1867, with a private score of 0.39968. My public leaderboard position is 404, with a public score of 0.42470.\n\nI am attaching three screenshots that explain the situation better than words can.\n\nThe first screenshot shows my submission history. Six submissions from within the deadline period each display \"Succeeded\" alongside a warning icon. All six have both public and private scores computed. Their private scores range from 0.502 to 0.544. The first place team on the private leaderboard finished at 0.49669, so any of these would have placed above the winner. The highest, at 0.54468, is above first place by approximately 0.048.\n\nThe second and third screenshots show my private and public leaderboard positions, with the URL visible for identification. The public screenshot also shows Kaggle congratulating me on 0.42470 as my best entry. That score came from two notebook reruns Kaggle executed on March 31, which are the only submissions marked COMPLETE in my history. Those are what went into the final evaluation.\n\nI am not writing to contest my placement or ask for anything to be reopened. The competition is closed and I understand the results are final. What I would like to understand is what the warning icon means on those six submissions. They were submitted before the deadline, Kaggle computed private scores for them, yet they were not included in the final evaluation.\n**\nThe reason this matters to me beyond the leaderboard is that this is real structural biology work. If those predictions were evaluated against the test set and represent a genuine quality difference, the methodology might be scientifically useful to the Stanford team independently of the competition outcome. The approach involved multi-template alignment with topological diversity selection across 5,716 competition templates and 167K Ribonanza structures, with constraint refinement passes for bond geometry and steric clashes.**\n\nThree questions if someone can help:\n\n1. What does the warning icon indicate on submissions that show both Succeeded and a computed private score?\n2. Were those scored predictions evaluated against the full private test set or only partially?\n3. Is there a way to share the methodology with the Stanford research team for scientific consideration, independent of the competition results?\n\nKaggle Suport shared that \"Usually that refers to an error in the submission.\". However, my submissions are scored using both set the public and the hidden (private). \n\nI am happy to share the relevant kernel and technical details privately.\n\nThank you,\n\nSantiago Maniches\nhttps://www.kaggle.com/santmani", "votes": -1 } ], "comments": [ { "id": 3455654, "author_name": "hongan", "author_url": "", "post_date": "2026-05-10T00:52:55.480000", "content": "You selected two submissions. After the competition closed, all submissions are marked by the warning icon to indicate that they are pre-deadline submissions, and those private scores you see there are computed using a placeholder private set (which is different from the final private test set, so they mean nothing in terms of final placement). Then, the two selected submissions were rerun using the new test data. so you should see two green submissions on top of the submissions history (rerun) which means the results of them are used to determine your final standing (public score and private score you now see on LB). Hope that helps! I was confused in the beginning as well
", "votes": 1, "replies": [ { "id": 3455666, "author_name": "Santiago Maniches", "author_url": "", "post_date": "2026-05-10T02:45:38.107000", "content": "Thank you for your response. Got it ! I thought these were being scored against the entire public and private data sets, and if that was the case (considering the scoring), I wanted to share the notebook with the organizers. Thanks again for the clarification!
", "votes": 0, "replies": [] } ] }, { "id": 3455040, "author_name": "Santiago Maniches", "author_url": "", "post_date": "2026-05-08T13:26:11.687000", "content": "Hi @sohiermse , @ryanholbrook and @wcukierski
\nSince the competition closed a while ago, I am tagging some members of Kaggle team, to see if this situation is worth further investigation. Thanks in advance for your time and sorry for tagging you directly, I am hoping the message to be discovered. Thank you again and sorry for the inconvenience!
", "votes": 0, "replies": [ { "id": 3459945, "author_name": "inversion", "author_url": "", "post_date": "2026-05-18T13:23:56.750000", "content": "Sorry for the delay. The response by @honganzhu is correct. Generally, we add a note saying why submissions were invalidated (e.g., due to a rerun), but that step was missed, which added to the confusion. We've updated some process steps to make things more clear in the future.
", "votes": 1, "replies": [ { "id": 3459950, "author_name": "Santiago Maniches", "author_url": "", "post_date": "2026-05-18T13:28:57.897000", "content": "That's no problem. It makes sense now, thank you so much for the clarification.
", "votes": 0, "replies": [] } ] } ] } ], "raw_markdown_by_id": { "3455654": "You selected two submissions. After the competition closed, all submissions are marked by the warning icon to indicate that they are pre-deadline submissions, and those private scores you see there are computed using a placeholder private set (which is different from the final private test set, so they mean nothing in terms of final placement). Then, the two selected submissions were rerun using the new test data. so you should see two green submissions on top of the submissions history (rerun) which means the results of them are used to determine your final standing (public score and private score you now see on LB). Hope that helps! I was confused in the beginning as well", "3455040": "Hi @sohiermse , @ryanholbrook and @wcukierski \n\nSince the competition closed a while ago, I am tagging some members of Kaggle team, to see if this situation is worth further investigation. Thanks in advance for your time and sorry for tagging you directly, I am hoping the message to be discovered. Thank you again and sorry for the inconvenience!", "3455038": "Hi,\n\nI competed as a solo participant under the team name \"Santiago Maniches.\" My final private leaderboard position is 504 out of 1867, with a private score of 0.39968. My public leaderboard position is 404, with a public score of 0.42470.\n\nI am attaching three screenshots that explain the situation better than words can.\n\nThe first screenshot shows my submission history. Six submissions from within the deadline period each display \"Succeeded\" alongside a warning icon. All six have both public and private scores computed. Their private scores range from 0.502 to 0.544. The first place team on the private leaderboard finished at 0.49669, so any of these would have placed above the winner. The highest, at 0.54468, is above first place by approximately 0.048.\n\nThe second and third screenshots show my private and public leaderboard positions, with the URL visible for identification. The public screenshot also shows Kaggle congratulating me on 0.42470 as my best entry. That score came from two notebook reruns Kaggle executed on March 31, which are the only submissions marked COMPLETE in my history. Those are what went into the final evaluation.\n\nI am not writing to contest my placement or ask for anything to be reopened. The competition is closed and I understand the results are final. What I would like to understand is what the warning icon means on those six submissions. They were submitted before the deadline, Kaggle computed private scores for them, yet they were not included in the final evaluation.\n**\nThe reason this matters to me beyond the leaderboard is that this is real structural biology work. If those predictions were evaluated against the test set and represent a genuine quality difference, the methodology might be scientifically useful to the Stanford team independently of the competition outcome. The approach involved multi-template alignment with topological diversity selection across 5,716 competition templates and 167K Ribonanza structures, with constraint refinement passes for bond geometry and steric clashes.**\n\nThree questions if someone can help:\n\n1. What does the warning icon indicate on submissions that show both Succeeded and a computed private score?\n2. Were those scored predictions evaluated against the full private test set or only partially?\n3. Is there a way to share the methodology with the Stanford research team for scientific consideration, independent of the competition results?\n\nKaggle Suport shared that \"Usually that refers to an error in the submission.\". However, my submissions are scored using both set the public and the hidden (private). \n\nI am happy to share the relevant kernel and technical details privately.\n\nThank you,\n\nSantiago Maniches\nhttps://www.kaggle.com/santmani" } }