Testing the Deliteralization Hypothesis in Human and Machine Translation

ArXi:2605.25686v1 Announce Type: new The recent shift from dedicated NMT systems to general-purpose LLMs has reshaped machine translation, with LLMs reported to produce fluent, less literal output than their predecessors. We test whether this shift extends to the deliteralization hypothesis, the long-standing claim from translation studies that translations become progressively less literal as they are drafted and revised.