conftrace_

token dropout

5 papers

Explore in graph

Co-occurring keywords

neural machine translation (2310) self-supervised learning (3777) vision transformer (1106) representation learning (6206) sequence compression (9) attention weight (68) scene context (24) fact memorization (6) language model pretraining (105) spatiotemporal representation (20)

Papers

Enhancing Elusive Clues in Knowledge Learning by Contrasting Attention of Language Models AAAI 2025

Accelerating Augmentation Invariance Pretraining NIPS 2024

Efficient Video Action Detection with Token Dropout and Context Refinement ICCV 2023

Neural Machine Translation without Embeddings NAACL 2021

Token Drop mechanism for Neural Machine Translation COLING 2020