Research output is growing faster than the systems used to evaluate it. The Predicting Replicability Challenge invites teams to use automated approaches to predict whether research findings will hold up when replicated—testing whether these methods can help assess research claims at scale. In Round 2, all teams outperformed the baseline model—an improvement over Round 1.

Read about Round 2 results & how to participate in the final round, which launches April 22!
https://www.cos.io/blog/predicting-replicability-challenge-round-2-results