ControllableGPT: A Ground-Up Designed Controllable GPT for Molecule Optimization
Large Language Models (LLMs) employ three popular training approaches: Masked Language Models (MLM), Causal Language Models (CLM), and Sequence-to-Sequence Models (seq2seq). However, each approach has its strengths and limitations, and faces challenges in addressing specific tasks that require contr...
Zapisane w:
| Główni autorzy: | , , , |
|---|---|
| Format: | Artykuł |
| Wydane: |
2025
|
| Hasła przedmiotowe: | |
| Dostęp online: | https://arxiv.org/abs/2502.10631 http://arxiv.org/abs/2502.10631 |
| Etykiety: |
Nie ma etykietki, Dołącz pierwszą etykiete!
|