| 1 |
Pathologies of Pre-trained Language Models in Few-shot Fine-tuning ...
|
|
|
|
Abstract:
Although adapting pre-trained language models with few examples has shown promising performance on text classification, there is a lack of understanding of where the performance gain comes from. In this work, we propose to answer this question by interpreting the adaptation behavior using post-hoc explanations from model predictions. By modeling feature statistics of explanations, we discover that (1) without fine-tuning, pre-trained models (e.g. BERT and RoBERTa) show strong prediction bias across labels; (2) although few-shot fine-tuning can mitigate the prediction bias and demonstrate promising prediction performance, our analysis shows models gain performance improvement by capturing non-task-related features (e.g. stop words) or shallow data patterns (e.g. lexical overlaps). These observations alert that pursuing model performance with fewer examples may incur pathological prediction behavior, which requires further sanity check on model predictions and careful design in model evaluations in few-shot ... : ACL 2022 Workshop on Insights from Negative Results in NLP ...
|
|
Keyword:
Computation and Language cs.CL; FOS Computer and information sciences
|
|
URL: https://dx.doi.org/10.48550/arxiv.2204.08039
https://arxiv.org/abs/2204.08039
|
|
BASE
|
|
Hide details
|
|
|
| 2 |
MetaXL: Meta Representation Transformation for Low-resource Cross-lingual Learning ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 3 |
A Dataset and Baselines for Multilingual Reply Suggestion ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 4 |
A Conditional Generative Matching Model for Multi-lingual Reply Suggestion ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 5 |
A Conditional Generative Matching Model for Multi-lingual Reply Suggestion ...
|
|
|
|
BASE
|
|
Show details
|
|
|
|
|
