Search Results for author:
Found 2 papers, 0 papers with code
Decoding Probing: Revealing Internal Linguistic Structures in Neural Language Models using Minimal Pairs
4) For Transformer-based models, both embeddings and attentions capture grammatical features but show distinct patterns.
Do self-supervised speech and language models extract similar representations as human brain?
Speech and language models trained through self-supervised learning (SSL) demonstrate strong alignment with brain activity during speech and language perception.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
