LLMs perform near human-level on NLI for familiar phrasal constructions, but drop by over 40% when asked to distinguish syntactically identical constructions with different meanings.
Neural reality of argument structure constructions
3 Pith papers cite this work, alongside 16 external citations. Polarity classification is still indexing.
3
Pith papers citing it
16
external citations · Crossref
fields
cs.CL 3representative citing papers
Transformers on synthetic grammar acquire abstract global statistical knowledge first, then local dependencies, showing initial over-generalizations that are later constrained.
Larger LLMs reproduce constructional productivity via entrenchment in coercion cases with nonce words but fail to use statistical preemption to avoid overgeneralizing semantically plausible but unobserved patterns.
citing papers explorer
-
Beyond Memorization: Assessing Semantic Generalization in Large Language Models Using Phrasal Constructions
LLMs perform near human-level on NLI for familiar phrasal constructions, but drop by over 40% when asked to distinguish syntactically identical constructions with different meanings.
-
Developmental approach reveals the statistical learning of Neural Language Models: Transformers generalize from the most abstract statistical patterns
Transformers on synthetic grammar acquire abstract global statistical knowledge first, then local dependencies, showing initial over-generalizations that are later constrained.
-
Linguistic Productivity in Large Language Models: Models Coerce, but do not Preempt
Larger LLMs reproduce constructional productivity via entrenchment in coercion cases with nonce words but fail to use statistical preemption to avoid overgeneralizing semantically plausible but unobserved patterns.