Tag
This paper argues that computer science conferences should require nonrepudiable experimental results to prevent tampering and denial, and introduces K-Veritas, a reference implementation for signed reports without accessing training data.