Search papers, labs, and topics across Lattice.
University of Tennessee
6
2
4
1
The rise of AI in software development is rendering traditional software measurement assumptions obsolete, necessitating a radical rethink of how we validate our findings.
Overcoming the challenge of over-merging, this methodology achieves a 99% AUC in identity resolution while drastically reducing mega-cluster sizes in the World of Code dataset.
Clumping errors in author identity mapping can lead to a staggering misrepresentation of developer contributions, with previous maps inflating precision metrics by failing to account for conflated identities.
Over 1 billion git commits are now classified by their identity trust tiers, revealing a significant shift towards cryptographically attested contributions in software development.
File-level copying in open source obscures vital dependency signals, leading to significant security and compliance risks that are often invisible to current dependency scanners.
Meta's RADAR system proves you can automate code review at scale for AI-generated code, slashing review times by 330% while *also* dramatically reducing revert and production incident rates.
