Pavel Stepachev
Data | Evals | Pre/Post-training
🔥 Updates:
- New preprint Tokenization: A Survey for Modern NLP
- Released tenderness — a library for text rendering and synthetic document generation.
- Paper accepted at ACL 2026 (main)
- Paper accepted at LREC 2026 (main)
- New release HPLT v4.0 aka The HPLT pool
- New preprint about tokenization