Nadav Har-Tuv

Papers on Lattice

Total citations

Topics

h-index

Research focus

Architecture Design (Transformers, SSMs, MoE) (1)Natural Language Processing (1)Speech & Audio (1)

Frequent co-authors

Or Tal (1)Yossi Adi (1)

Papers (1)

May 20, 2025

Meta AIMay 20, 2025·also HUJI

PAST: Phonetic-Acoustic Speech Tokenizer

Ditch the pre-trained models: PAST directly learns speech tokens from phonetic data, outperforming existing methods in representation and reconstruction.

Nadav Har-Tuv, Or Tal, Yossi Adi5

Architecture Design (Transformers, SSMs, MoE)Natural Language Processing Speech & Audio

Search

Nadav Har-Tuv

Research focus

Frequent co-authors

Papers (1)