A linear projection of Word2Vec and GloVe embeddings onto eight part-of-speech axes yields compact interpretable vectors, and combining them with the original vectors gives small gains on a few downstream tasks.
Each pair of data in all of these datasets are annotated a human generated similarity score
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Interpretable Syntactic Representations Enable Hierarchical Word Vectors
A linear projection of Word2Vec and GloVe embeddings onto eight part-of-speech axes yields compact interpretable vectors, and combining them with the original vectors gives small gains on a few downstream tasks.