Comparative Evaluation of Pretrained Transfer Learning Models on Automatic Short Answer Grading

ELMo outperformed the other three models, BERT, GPT, and GPT-2 . We train with a single feature, cosine similarity, extracted from the embeddings of these models . We compare the RMSE scores and correlation measurements of the four models with previous works on Mohler dataset . We also, briefly describe the four transfer learning models and conclude with the possible causes of poor results of the models . The authors conclude that the possible cause of the poor results is the failure to use the same techniques as those used in previous works . The results are published in the journal Nature Nature, Springer Springer, on October 31, 2013 . The author of this article is entitled, “

Links: PDF - Abstract

Code :


Keywords : models - results - learning - nature - transfer -

Leave a Reply

Your email address will not be published. Required fields are marked *