TPDM: Selectively Removing Positional Information for Zero-shot Translation via Token-Level Position Disentangle Module

TL;DR

This paper introduces TPDM, a novel framework that selectively removes positional information at the token level to enhance zero-shot translation in multilingual neural machine translation, achieving significant improvements.

Contribution

It proposes a token-level position disentangle module (TPDM) that learns to preserve useful positional information selectively, addressing limitations of previous methods.

Findings

Significant improvement in zero-shot translation performance.

Reduced performance loss in supervised translation.

Effective disentanglement of positional information at token level.

Abstract

Due to Multilingual Neural Machine Translation's (MNMT) capability of zero-shot translation, many works have been carried out to fully exploit the potential of MNMT in zero-shot translation. It is often hypothesized that positional information may hinder the MNMT from outputting a robust encoded representation for decoding. However, previous approaches treat all the positional information equally and thus are unable to selectively remove certain positional information. In sharp contrast, this paper investigates how to learn to selectively preserve useful positional information. We describe the specific mechanism of positional information influencing MNMT from the perspective of linguistics at the token level. We design a token-level position disentangle module (TPDM) framework to disentangle positional information at the token level based on the explanation. Our experiments demonstrate that our framework improves zero-shot translation by a large margin while reducing the performance loss in the supervised direction compared to previous works.