Lukas Edman
leukas
AI & ML interests
NLP, tokenization, pretraining, low-resource
Recent Activity
liked a Space 8 days ago
BabyLM-community/BabyLM-Leaderboard-2026 new activity 15 days ago
leukas/Qwenfood:[bot] Conversion to Parquet published a model 16 days ago
leukas/Qwenfant-0.4BOrganizations
Are Character-level Translations Worth the Wait?
Collection of trained models for the paper: Are Character-level Translations Worth the Wait?
BabyLM 2023
Final model for BabyLM 2023
BabyLM 2026
BabyLM 2025
Are Character-level Translations Worth the Wait?
Collection of trained models for the paper: Are Character-level Translations Worth the Wait?
BabyLM 2024
BabyLM2024 models
BabyLM 2023
Final model for BabyLM 2023
CUTE
The CUTE benchmark is an LLM benchmark, testing LLMs' understanding of orthography. Check out our github here: https://github.com/Leukas/cute