BERTs of a feather do not generalize together: Large variability in generalization across models with similar test set performance

11/07/2019

∙

If the same neural architecture is trained multiple times on the same dataset, will it make similar linguistic generalizations across runs? To study this question, we fine-tuned 100 instances of BERT on the Multi-genre Natural Language Inference (MNLI) dataset and evaluated them on the HANS dataset, which measures syntactic generalization in natural language inference. On the MNLI development set, the behavior of all instances was remarkably consistent, with accuracy ranging between 83.6 varied widely in their generalization performance. For example, on the simple case of subject-object swap (e.g., knowing that "the doctor visited the lawyer" does not entail "the lawyer visited the doctor"), accuracy ranged from 0.00 66.2 are equally attractive to a low-bias learner such as a neural network; decreasing the variability may therefore require models with stronger inductive biases.

READ FULL TEXT

BERTs of a feather do not generalize together: Large variability in generalization across models with similar test set performance

Testing the Generalization Power of Neural Network Models Across NLI Benchmarks

Does syntax need to grow on trees? Sources of hierarchical inductive bias in sequence-to-sequence networks

Syntactic Data Augmentation Increases Robustness to Inference Heuristics

Probing Linguistic Systematicity

Robust Natural Language Inference Models with Example Forgetting

ExpBERT: Representation Engineering with Natural Language Explanations

Individual predictions matter: Assessing the effect of data ordering in training fine-tuned CNNs for medical imaging

BERTs of a feather do not generalize together: Large variability in generalization across models with similar test set performance

Related Research

Testing the Generalization Power of Neural Network Models Across NLI Benchmarks

Does syntax need to grow on trees? Sources of hierarchical inductive bias in sequence-to-sequence networks

Syntactic Data Augmentation Increases Robustness to Inference Heuristics

Probing Linguistic Systematicity

Robust Natural Language Inference Models with Example Forgetting

ExpBERT: Representation Engineering with Natural Language Explanations

Individual predictions matter: Assessing the effect of data ordering in training fine-tuned CNNs for medical imaging