Hi team,
I noticed that the recent updates to the coding data haven’t been publicly available since November 2024. I understand the concerns about contamination, but is there a way to access those questions to better understand why some rankings have changed on the leaderboard and if it is contributed to the data distribution or difficulty of questions in the new dataset? I’d also be curious to run some of the latest models on both the previous questions and the newest ones to observe any trends.