Announcement_3
Our paper, When Does Generating More Help? Disentangling Fixed-Source Synthesis from Source Expansion in Synthetic Data Scaling, has been accepted to Findings of EMNLP 2026.
Our paper, When Does Generating More Help? Disentangling Fixed-Source Synthesis from Source Expansion in Synthetic Data Scaling, has been accepted to Findings of EMNLP 2026.