For English-German numerical localisation with five small LLMs, in-context learning with explicit formatting rules beats direct translation, chain-of-thought, and post-editing in mean accuracy.
Findings of the 2020 Conference on Machine Translation (WMT20)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Analysis of Numerical Localisation in LLM Translations
For English-German numerical localisation with five small LLMs, in-context learning with explicit formatting rules beats direct translation, chain-of-thought, and post-editing in mean accuracy.