Self-Supervised Speech Representations are More Phonetic than Semantic

Re-assign community