paper-with-me

홈 › Papers

RetroGAN: A Cyclic Post-Specialization System for Improving Out-of-Knowledge and Rare Word Representations

2021-08-30 · Findings (ACL) 2021 8 · Pedro Colon-Hernandez, Yida Xin, Henry Lieberman, Catherine Havasi, Cynthia Breazeal, Peter Chin

Retrofitting is a technique used to move word vectors closer together or further apart in their space to reflect their relationships in a Knowledge Base (KB). However, retrofitting only works on concepts that are present in that KB. RetroGAN uses a pair of Generative Adversarial Networks (GANs) to learn a one-to-one mapping between concepts and their retrofitted counterparts. It applies that mapping (post-specializes) to handle concepts that do not appear in the original KB in a manner similar to how some natural language systems handle out-of-vocabulary entries. We test our system on three word-similarity benchmarks and a downstream sentence simplification task and achieve the state of the art (CARD-660). Altogether, our results demonstrate our system's effectiveness for out-of-knowledge and rare word generalization.

📄 PDF Abstract BibTeX arXiv:2108.12941

Code (1)

pedrocolon93/retrogan 공식 구현 pytorch

Tasks

SentenceWord Similarity

Similar Papers 제목 키워드 기반

Specializing Distributional Vectors of All Words for Lexical Entailment

2019-08-01 · WS 2019 8 · Aishwarya Kamath, Jonas Pfeiffer, Edoardo Maria Ponti, Goran Glava{\v{s}} 외

Semantic specialization methods fine-tune distributional word vectors using lexical knowledge from external resources (e.g. WordNet) to accentuate a particular relation between words. However, such post-processing method…

AllCross-Lingual TransferLexical EntailmentRelation+2

Adversarial Propagation and Zero-Shot Cross-Lingual Transfer of Word Vector Specialization

2018-09-11 · EMNLP 2018 10 · Edoardo Maria Ponti, Ivan Vulić, Goran Glavaš, Nikola Mrkšić 외

Semantic specialization is the process of fine-tuning pre-trained distributional word vectors using external lexical knowledge (e.g., WordNet) to accentuate a particular semantic relation in the specialized vector space.…

Cross-Lingual Transferdialog state trackingLexical SimplificationWord Similarity+1

Enhancing Word Embeddings with Knowledge Extracted from Lexical Resources

2020-05-20 · ACL 2020 6 · Magdalena Biesialska, Bardia Rafieian, Marta R. Costa-jussà

In this work, we present an effective method for semantic specialization of word vector representations. To this end, we use traditional word embeddings and apply specialization methods to better capture semantic relatio…

dialog state trackingWord EmbeddingsWord Similarity

DBES: A Systematic Benchmark and Metric Suite for Evaluating Expert Specialization in Large-Scale MoEs

2026-05-18 · Jing Wang, Hongxuan Lu, Jazze Young, Shu Wang 외 arxiv

Expert specialization in Mixture-of-Experts (MoE) models remains poorly understood, with traditional evaluations conflating architectural load-balancing with functional specialization. We introduce DBES, a comprehensive …

Implicit Model Specialization through DAG-based Decentralized Federated Learning

2021-11-01 · Jossekin Beilharz, Bjarne Pfitzner, Robert Schmid, Paul Geppert 외

Federated learning allows a group of distributed clients to train a common machine learning model on private data. The exchange of model updates is managed either by a central entity or in a decentralized way, e.g. by a …

Federated Learning