Severity-aware loss weighting improves Arabic medical text generation by up to 12.10% across ten LLMs by scaling token loss according to automatically derived clinical severity probabilities.
Springer Nature Singapore, Singapore (2025)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Severity-Aware Weighted Loss for Arabic Medical Text Generation
Severity-aware loss weighting improves Arabic medical text generation by up to 12.10% across ten LLMs by scaling token loss according to automatically derived clinical severity probabilities.