A Systematic Assessment of Syntactic Generalization in Neural Language Models
This work addresses the problem of understanding whether optimizing for predictive performance leads to human-like syntactic knowledge in language models, which is important for researchers in natural language processing and AI, though it is incremental as it builds on existing evaluation frameworks.
The study systematically evaluated syntactic generalization in neural language models, finding that model architecture has a greater impact on performance than dataset size, with sequential models underperforming others, and revealing a dissociation between perplexity and syntactic generalization.
While state-of-the-art neural network models continue to achieve lower perplexity scores on language modeling benchmarks, it remains unknown whether optimizing for broad-coverage predictive performance leads to human-like syntactic knowledge. Furthermore, existing work has not provided a clear picture about the model properties required to produce proper syntactic generalizations. We present a systematic evaluation of the syntactic knowledge of neural language models, testing 20 combinations of model types and data sizes on a set of 34 English-language syntactic test suites. We find substantial differences in syntactic generalization performance by model architecture, with sequential models underperforming other architectures. Factorially manipulating model architecture and training dataset size (1M--40M words), we find that variability in syntactic generalization performance is substantially greater by architecture than by dataset size for the corpora tested in our experiments. Our results also reveal a dissociation between perplexity and syntactic generalization performance.