M. Gao

Only 5.4% of coding agent attempts successfully complete a migration while preserving behavioral correctness, exposing a critical gap in current AI capabilities for software evolution.

Agents struggle to significantly improve training algorithms, with the best only achieving 25% of the potential optimization gap.

AdvFD not only boosts visual quality in generative models but also tackles the pitfalls of static feature spaces that lead to stagnation in performance.

Switching music tokenization can halve the Frechet Music Distance, proving that representation trumps model size in text-to-music generation.