Computer Science Conferences Should Require Nonrepudiable Experimental Results

A position paper argues computer science conferences should require tamper-evident proof that published numbers came from real executed code.

The proposal centers on nonrepudiable attestations, making experimental results harder to falsify or later dispute.

The paper introduces K-Veritas, a Go-based reference prototype, to show how such verification could work in practice.

It frames this as a stronger standard for reproducible research and better experimental integrity.