Scilit logo
  • Rankings
  • Search Publications
    Cited-by Search
  • Sources
  • Publishers
  • Scholars
    Top Cited Scholars
  • Organizations
  • Scilit
    About Us
    Advisory Board
    Data Sources
    Career
    Support / Contact
    User Guide
    Publisher User
    Search Function
    Rankings Function
    API Resource
    Scilit Knowledge
    Metrics
    Scilit Recognitions
    Publications Classification
    Media
    News
    Blog
Register
  1. Home
  2. Publications
  3. A Contrastive Self-distillation BERT with Kernel Alignment-Based Inference
  1. Home
  2. Publications
  3. A Contrastive Self-distillation BERT with Kernel Alignment-Based Inference

A Contrastive Self-distillation BERT with Kernel Alignment-Based Inference

Authors:
  • YXYangyan Xu
    Yangyan Xu
  • FYFangfang Yuan
    Fangfang Yuan
  • CCCong Cao
    Cong Cao
  • MSMajing Su
    Majing Su
  • YLYuhai Lu
    Yuhai Lu
  • YLYanbing Liu
    Yanbing Liu

Publisher website
Google Scholar
  • 26 June 2023
    • book chapter
    • Published by Springer Nature
    • p. 553-565
    • DOI:10.1007/978-3-031-35995-8_39
Abstract
No abstract available
Keywords
  • EARLY EXIT
  • CONTRASTIVE LEARNING
  • SELF-DISTILLATION
  • CENTERED KERNEL ALIGNMENT

This publication has 20 references indexed in Scilit:

  • Accelerating Pretrained Language Model Inference Using Weighted Ensemble Self-distillation
    2021
  • LeeBERT: Learned Early Exit for BERT with cross-level optimization
    2021
  • BERT-of-Theseus: Compressing BERT by Progressive Module Replacing
    2020
  • Transformers: State-of-the-Art Natural Language Processing
    2020
  • The Right Tool for the Job: Matching Model and Instance Complexities
    2020
  • DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference
    2020
  • FastBERT: a Self-distilling BERT with Adaptive Inference Time
    2020
  • Patient Knowledge Distillation for BERT Model Compression
    2019
  • GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding
    2018
  • BranchyNet: Fast inference via early exiting from deep neural networks
    2016

Publications with shared reference occurs when two publications share one or more common references. This publication has 15063 Publications with shared reference indexed in Scilit, top 10 records with the most shared references are listed below:

View publications in search results
Cited by 3 articles
Scilit logo

Have questions? Email us at support@scilit.com

© 2026 MDPI (Basel, Switzerland) unless otherwise stated.

MDPI initiatives

  • MDPI Journals
  • MDPI Books
  • Sciforum
  • Proceedings Series
  • Preprints.org
  • SciProfiles
  • Encyclopedia
  • JAMS

Legal

  • Privacy
  • Terms
  • About
  • Contact
  • Disclaimer

Follow us

  • facebook icon
  • x icon
  • linkedin icon

Join us

  • Career

© 2026 MDPI (Basel, Switzerland) unless otherwise stated.