Hi everyone,
\nWe’ve just updated the evaluation dataset for this competition to improve scoring accuracy.
\nAs a result, the number of required rows in each submission has changed.
What Changed:
\nThe evaluation dataset has been updated with the final version (evaluation-set-30s).
\nThe new dataset contains 64,208 tokens.
\nYour submission CSV must now include exactly 64,208 rows — one for each mint_token_id in the evaluation set.
\nWhat You Need to Do
\nPlease re-generate your predictions using the updated dataset.
\nRe-upload your submission file (same format):
\nmint_token_id,is_target
\n,0
\n,1
\n…
Old submissions (from before this update) will no longer be valid and will show an error.
\nDeadline and Leaderboard
\nThe competition timeline remains unchanged.
\nThe leaderboard will be refreshed automatically once new submissions are processed.
\nThank you for your patience — this update ensures fairer and more accurate evaluation for everyone.
\nIf you have any questions or run into issues generating the new submission, please ask in this thread.
— Twain
", "messages": [ { "id": 3320100, "postDate": "2025-11-12T08:36:05.863Z", "content": "Hi Twain,\nI hope you’re doing well. I noticed that the competition deadline has passed, but the leaderboard and final results haven’t been announced yet.\nCould you please let us know if there’s an estimated date for the announcement or if any review process is still ongoing?
\nThanks a lot for your time and all the work organizing this competition!
", "rawMarkdown": "Hi Twain,\nI hope you’re doing well. I noticed that the competition deadline has passed, but the leaderboard and final results haven’t been announced yet.\nCould you please let us know if there’s an estimated date for the announcement or if any review process is still ongoing?\n\nThanks a lot for your time and all the work organizing this competition!", "votes": 1 }, { "id": 3316915, "postDate": "2025-11-10T21:10:09.793Z", "content": "Hello, @twainayar could you confirm whether submission with recall <75% recall will be discarded from private leaderboard, so that we can select the appropriate submissions ?\nThank you
", "rawMarkdown": "Hello, @twainayar could you confirm whether submission with recall <75% recall will be discarded from private leaderboard, so that we can select the appropriate submissions ?\nThank you", "votes": 1 }, { "id": 3301567, "postDate": "2025-10-13T16:46:16.743Z", "content": "Hi everyone,
\nWe’ve just updated the evaluation dataset for this competition to improve scoring accuracy.
\nAs a result, the number of required rows in each submission has changed.
What Changed:
\nThe evaluation dataset has been updated with the final version (evaluation-set-30s).
\nThe new dataset contains 64,208 tokens.
\nYour submission CSV must now include exactly 64,208 rows — one for each mint_token_id in the evaluation set.
\nWhat You Need to Do
\nPlease re-generate your predictions using the updated dataset.
\nRe-upload your submission file (same format):
\nmint_token_id,is_target
\n,0
\n,1
\n…
Old submissions (from before this update) will no longer be valid and will show an error.
\nDeadline and Leaderboard
\nThe competition timeline remains unchanged.
\nThe leaderboard will be refreshed automatically once new submissions are processed.
\nThank you for your patience — this update ensures fairer and more accurate evaluation for everyone.
\nIf you have any questions or run into issues generating the new submission, please ask in this thread.
— Twain
", "rawMarkdown": "Hi everyone,\n\nWe’ve just updated the evaluation dataset for this competition to improve scoring accuracy.\nAs a result, the number of required rows in each submission has changed.\n\nWhat Changed:\n\nThe evaluation dataset has been updated with the final version (evaluation-set-30s).\n\nThe new dataset contains 64,208 tokens.\n\nYour submission CSV must now include exactly 64,208 rows — one for each mint_token_id in the evaluation set.\n\nWhat You Need to Do\n\nPlease re-generate your predictions using the updated dataset.\n\nRe-upload your submission file (same format):\n\nmint_token_id,is_target\nHello, should only the CatBoost model be used, or is the model choice flexible?
", "rawMarkdown": "Hello, should only the CatBoost model be used, or is the model choice flexible?", "replies": [ { "id": 3310509, "postDate": "2025-11-03T05:50:16.557Z", "content": "Only CatBoost
", "rawMarkdown": "Only CatBoost" } ] }, { "id": 3302088, "postDate": "2025-10-15T01:41:12.627Z", "content": "Hello, is it possible to upload the target token to Kaggle as well?
\nBecause of the internet restrictions, some people may not have access to any google products.
\nThanks.
https://www.kaggle.com/datasets/zinebrafik/target/data
", "rawMarkdown": "**[https://www.kaggle.com/datasets/zinebrafik/target/data](https://www.kaggle.com/datasets/zinebrafik/target/data)**", "isDeleted": true } ] }, { "id": 3301710, "postDate": "2025-10-14T02:43:36.123Z", "content": "Hi, I have submitted with new dataset, but the old submission seems to be still valid, so it overwrite the new one on the leaderboard. Anyway to get rid of the previous one?
", "rawMarkdown": "Hi, I have submitted with new dataset, but the old submission seems to be still valid, so it overwrite the new one on the leaderboard. Anyway to get rid of the previous one?", "replies": [ { "id": 3302071, "postDate": "2025-10-14T23:31:09.710Z", "content": "Sr, they become invalid now
", "rawMarkdown": "Sr, they become invalid now" } ] }, { "id": 3301621, "postDate": "2025-10-13T19:23:26.610Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 3320100, "author_name": "Martin Szerment", "author_url": "", "post_date": "2025-11-12T08:36:05.863000", "content": "Hi Twain,\nI hope you’re doing well. I noticed that the competition deadline has passed, but the leaderboard and final results haven’t been announced yet.\nCould you please let us know if there’s an estimated date for the announcement or if any review process is still ongoing?
\nThanks a lot for your time and all the work organizing this competition!
", "votes": 1, "replies": [] }, { "id": 3316915, "author_name": "juliencst", "author_url": "", "post_date": "2025-11-10T21:10:09.793000", "content": "Hello, @twainayar could you confirm whether submission with recall <75% recall will be discarded from private leaderboard, so that we can select the appropriate submissions ?\nThank you
", "votes": 1, "replies": [] }, { "id": 3309976, "author_name": "Cyrus", "author_url": "", "post_date": "2025-11-01T19:40:02.980000", "content": "Hello, should only the CatBoost model be used, or is the model choice flexible?
", "votes": 0, "replies": [ { "id": 3310509, "author_name": "nunjunj", "author_url": "", "post_date": "2025-11-03T05:50:16.557000", "content": "Only CatBoost
", "votes": 0, "replies": [] } ] }, { "id": 3302088, "author_name": "zhenghua zhu", "author_url": "", "post_date": "2025-10-15T01:41:12.627000", "content": "Hello, is it possible to upload the target token to Kaggle as well?
\nBecause of the internet restrictions, some people may not have access to any google products.
\nThanks.
https://www.kaggle.com/datasets/zinebrafik/target/data
", "votes": 0, "replies": [] } ] }, { "id": 3301710, "author_name": "Lê Đức Tùng Dương ", "author_url": "", "post_date": "2025-10-14T02:43:36.123000", "content": "Hi, I have submitted with new dataset, but the old submission seems to be still valid, so it overwrite the new one on the leaderboard. Anyway to get rid of the previous one?
", "votes": 0, "replies": [ { "id": 3302071, "author_name": "Lê Đức Tùng Dương ", "author_url": "", "post_date": "2025-10-14T23:31:09.710000", "content": "Sr, they become invalid now
", "votes": 0, "replies": [] } ] }, { "id": 3301621, "author_name": "", "author_url": "", "post_date": "2025-10-13T19:23:26.610000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "3320100": "Hi Twain,\nI hope you’re doing well. I noticed that the competition deadline has passed, but the leaderboard and final results haven’t been announced yet.\nCould you please let us know if there’s an estimated date for the announcement or if any review process is still ongoing?\n\nThanks a lot for your time and all the work organizing this competition!", "3316915": "Hello, @twainayar could you confirm whether submission with recall <75% recall will be discarded from private leaderboard, so that we can select the appropriate submissions ?\nThank you", "3301567": "Hi everyone,\n\nWe’ve just updated the evaluation dataset for this competition to improve scoring accuracy.\nAs a result, the number of required rows in each submission has changed.\n\nWhat Changed:\n\nThe evaluation dataset has been updated with the final version (evaluation-set-30s).\n\nThe new dataset contains 64,208 tokens.\n\nYour submission CSV must now include exactly 64,208 rows — one for each mint_token_id in the evaluation set.\n\nWhat You Need to Do\n\nPlease re-generate your predictions using the updated dataset.\n\nRe-upload your submission file (same format):\n\nmint_token_id,is_target\n