Autoregressive transformers achieve 100% first-order exemplar retrieval but only chance-level second-order generalization to novel nouns on controlled synthetic word-learning tasks.
Learning overhypotheses with hierarchical Bayesian models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Exemplar Retrieval Without Overhypothesis Induction: Limits of Distributional Sequence Learning in Early Word Learning
Autoregressive transformers achieve 100% first-order exemplar retrieval but only chance-level second-order generalization to novel nouns on controlled synthetic word-learning tasks.