LLMs achieve high-quality translations of Galen’s expository Greek (MQM 95.2/100) but lower and bimodal quality on pharmacological texts (79.9/100), with terminology rarity (corpus frequency) predicting failure at r = -0.97.
The automated tests are cheap, fast, and widely standardized, but of uncertain efficacy
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Evaluating LLM-Based Translation of a Low-Resource Technical Language: The Medical and Philosophical Greek of Galen
LLMs achieve high-quality translations of Galen’s expository Greek (MQM 95.2/100) but lower and bimodal quality on pharmacological texts (79.9/100), with terminology rarity (corpus frequency) predicting failure at r = -0.97.