Indicatements that character language models learn English morpho-syntactic units and regularities
This addresses the problem of understanding model capabilities in natural language processing for researchers, showing incremental insights into morphological abstraction.
The study investigated whether character language models learn abstract morphological regularities in English, finding that they can identify word and morpheme boundaries and capture linguistic properties, including selectional restrictions of derivational morphemes.
Character language models have access to surface morphological patterns, but it is not clear whether or how they learn abstract morphological regularities. We instrument a character language model with several probes, finding that it can develop a specific unit to identify word boundaries and, by extension, morpheme boundaries, which allows it to capture linguistic properties and regularities of these units. Our language model proves surprisingly good at identifying the selectional restrictions of English derivational morphemes, a task that requires both morphological and syntactic awareness. Thus we conclude that, when morphemes overlap extensively with the words of a language, a character language model can perform morphological abstraction.