paper-with-me

홈 › Papers

Autoencoder-Based Framework to Capture Vocabulary Quality in NLP

2025-02-28 · Vu Minh Hoang Dang, Rakesh M. Verma

Linguistic richness is essential for advancing natural language processing (NLP), as dataset characteristics often directly influence model performance. However, traditional metrics such as Type-Token Ratio (TTR), Vocabulary Diversity (VOCD), and Measure of Lexical Text Diversity (MTLD) do not adequately capture contextual relationships, semantic richness, and structural complexity. In this paper, we introduce an autoencoder-based framework that uses neural network capacity as a proxy for vocabulary richness, diversity, and complexity, enabling a dynamic assessment of the interplay between vocabulary size, sentence structure, and contextual depth. We validate our approach on two distinct datasets: the DIFrauD dataset, which spans multiple domains of deceptive and fraudulent text, and the Project Gutenberg dataset, representing diverse languages, genres, and historical periods. Experimental results highlight the robustness and adaptability of our method, offering practical guidance for dataset curation and NLP model design. By enhancing traditional vocabulary evaluation, our work fosters the development of more context-aware, linguistically adaptive NLP systems.

📄 PDF Abstract BibTeX arXiv:2503.00209

Code (0)

등록된 구현이 없습니다.

Tasks

DiversitySentence

Similar Papers 제목 키워드 기반

Phonological Representation Learning for Isolated Signs Improves Out-of-Vocabulary Generalization

2025-09-05 · Lee Kezar, Zed Sehyr, Jesse Thomason arxiv

Sign language datasets are often not representative in terms of vocabulary, underscoring the need for models that generalize to unseen signs. Vector quantization is a promising approach for learning discrete, token-like …

Representation Learning

VASAE: Naming SAE Dictionary Directions with Vocabulary-Aligned Anchoring

2026-06-26 · Kairui Zhang, Ziwen Yu, Zahraa S. Abdallah, Martha Lewis arxiv

Sparse autoencoders (SAEs) provide useful decompositions of Transformer residual streams, but their learned features are usually named post hoc rather than directly connected to the Transformer's token vocabulary. We int…

Self-StrAE at SemEval-2024 Task 1: Making Self-Structuring AutoEncoders Learn More With Less

2024-04-02 · Mattia Opper, N. Siddharth

This paper presents two simple improvements to the Self-Structuring AutoEncoder (Self-StrAE). Firstly, we show that including reconstruction to the vocabulary as an auxiliary objective improves representation quality. Se…

PerTok: Expressive Encoding and Modeling of Symbolic Musical Ideas and Variations

2024-10-02 · Julian Lenz, Anirudh Mani

We introduce Cadenza, a new multi-stage generative framework for predicting expressive variations of symbolic musical ideas as well as unconditional generations. To accomplish this we propose a novel MIDI encoding method…

Decoder

LEGO-SLAM: Language-Embedded Gaussian Optimization SLAM

2025-11-20 · Sibaek Lee, Seongbo Ha, Kyeongsu Kang, Joonyeol Choi 외 arxiv

Recent advances in 3D Gaussian Splatting (3DGS) have enabled Simultaneous Localization and Mapping (SLAM) systems to build photorealistic maps. However, these maps lack the open-vocabulary semantic understanding required…