paper-with-me

홈 › Papers

Reconsidering Degeneration of Token Embeddings with Definitions for Encoder-based Pre-trained Language Models

2024-08-02 · Ying Zhang, Dongyuan Li, Manabu Okumura

Learning token embeddings based on token co-occurrence statistics has proven effective for both pre-training and fine-tuning in natural language processing. However, recent studies have pointed out that the distribution of learned embeddings degenerates into anisotropy (i.e., non-uniform distribution), and even pre-trained language models (PLMs) suffer from a loss of semantics-related information in embeddings for low-frequency tokens. This study first analyzes the fine-tuning dynamics of encoder-based PLMs and demonstrates their robustness against degeneration. On the basis of this analysis, we propose DefinitionEMB, a method that utilizes definitions to re-construct isotropically distributed and semantics-related token embeddings for encoder-based PLMs while maintaining original robustness during fine-tuning. Our experiments demonstrate the effectiveness of leveraging definitions from Wiktionary to re-construct such embeddings for two encoder-based PLMs: RoBERTa-base and BART-large. Furthermore, the re-constructed embeddings for low-frequency tokens improve the performance of these models across various GLUE and four text summarization datasets.

📄 PDF Abstract BibTeX arXiv:2408.01308

Code (0)

등록된 구현이 없습니다.

Tasks

Text Summarization

Similar Papers 제목 키워드 기반

Rare Tokens Degenerate All Tokens: Improving Neural Text Generation via Adaptive Gradient Gating for Rare Token Embeddings

2021-11-16 · ACL ARR November 2021 11 · Anonymous

Recent studies have determined that the learned token embeddings of large-scale neural language models are degenerated to be anisotropic with a narrow-cone shape. This phenomenon, called the representation degeneration p…

AllLanguage ModelingLanguage ModellingMachine Translation+2

Rare Tokens Degenerate All Tokens: Improving Neural Text Generation via Adaptive Gradient Gating for Rare Token Embeddings

2021-09-07 · ACL 2022 5 · Sangwon Yu, Jongyoon Song, Heeseung Kim, Seong-min Lee 외

Recent studies have determined that the learned token embeddings of large-scale neural language models are degenerated to be anisotropic with a narrow-cone shape. This phenomenon, called the representation degeneration p…

AllLanguage ModelingLanguage ModellingMachine Translation+3

Isotropic Contextual Representations through Variational Regularization

2021-09-29 · Cornelia Ferner, Stefan Wegenkittl

Contextual language representations achieve state-of-the-art performance across various natural language processing tasks. However, these representations have been shown to suffer from the degeneration problem, i.e. they…

DecoderSentence

Concept Tokens: Learning Behavioral Embeddings Through Concept Definitions

2026-01-08 · Ignacio Sastre, Aiala Rosá arxiv

We propose Concept Tokens, a lightweight method that adds a new special token to a pretrained LLM and learns only its embedding from multiple natural language definitions of a target concept, where occurrences of the con…

Question Answering

Learning Sense Embeddings from Definitions in Dictionaries

2022-01-16 · ACL ARR January 2022 1 · Anonymous

We introduce a method for learning to embed word senses as defined in a given set of given dictionaries. In our approach, sense definition pairs, <word, definition> are transformed into low-dimension vectors aimed at max…

Re-Ranking