LMMS Reloaded: Transformer-based Sense Embeddings for Disambiguation and Beyond

TL;DR

This paper introduces a novel, layer-informed method for generating sense embeddings from Transformer-based language models, improving performance on sense disambiguation and related tasks by leveraging all model layers.

Contribution

It presents a principled approach to utilize all layers of NLMs for sense embedding creation, outperforming prior methods and demonstrating versatility across multiple sense-related tasks.

Findings

Layer-informed sense embeddings outperform previous models.

Using all layers of NLMs enhances sense disambiguation accuracy.

Sense embeddings show versatility beyond WSD tasks.

Abstract

Distributional semantics based on neural approaches is a cornerstone of Natural Language Processing, with surprising connections to human meaning representation as well. Recent Transformer-based Language Models have proven capable of producing contextual word representations that reliably convey sense-specific information, simply as a product of self-supervision. Prior work has shown that these contextual representations can be used to accurately represent large sense inventories as sense embeddings, to the extent that a distance-based solution to Word Sense Disambiguation (WSD) tasks outperforms models trained specifically for the task. Still, there remains much to understand on how to use these Neural Language Models (NLMs) to produce sense embeddings that can better harness each NLM's meaning representation abilities. In this work we introduce a more principled approach to leverage information from all layers of NLMs, informed by a probing analysis on 14 NLM variants. We also emphasize the versatility of these sense embeddings in contrast to task-specific models, applying them on several sense-related tasks, besides WSD, while demonstrating improved performance using our proposed approach over prior work focused on sense embeddings. Finally, we discuss unexpected findings regarding layer and model performance variations, and potential applications for downstream tasks.