Do We Need Distinct Representations for Every Speech Token? Unveiling and Exploiting Redundancy in Large Speech Language Models.

Bajian Xiang, Tingwei Guo, Xuan Chen, Yang Han

Browse the full ACL paper archive.