Pavel Stepachev
Multilingual & Multimodal Data | LLM Pre/Mid-training
🔥 Updates:
- Released tenderness — a library for synthetic data 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
- Paper accepted at ACL 2025 (main)
- Paper accepted at EMNLP 2024 (WMT24, spotlight)
- Paper accepted at NAACL 2024 (AmericasNLP 2024)