paper-with-me

Papers

Difference-Masking: Choosing What to Mask in Continued Pretraining

2023-05-23 · Alex Wilf, Syeda Nahida Akter, Leena Mathur, Paul Pu Liang, Sheryl Mathew, Mengrou Shou, Eric Nyberg, Louis-Philippe Morency

The self-supervised objective of masking-and-predicting has led to promising performance gains on a variety of downstream tasks. However, while most approaches randomly mask tokens, there is strong intuition that deciding what to mask can substantially improve learning outcomes. We investigate this in continued pretraining setting in which pretrained models continue to pretrain on domain-specific data before performing some downstream task. We introduce Difference-Masking, a masking strategy that automatically chooses what to mask during continued pretraining by considering what makes a task domain different from the pretraining domain. Empirically, we find that Difference-Masking outperforms baselines on continued pretraining settings across four diverse language-only and multimodal video tasks.

📄 PDF Abstract BibTeX arXiv:2305.14577

Code (1)

abwilf/difference-masking 공식 구현 jax

Tasks

Self-Supervised Learning

Similar Papers 제목 키워드 기반

Self-supervised adversarial masking for 3D point cloud representation learning

2023-07-11 · Michał Szachniewicz, Wojciech Kozłowski, Michał Stypułkowski, Maciej Zięba

Self-supervised methods have been proven effective for learning deep representations of 3D point cloud data. Although recent methods in this domain often rely on random masking of inputs, the results of this approach can…

Representation Learning

Frequency Is What You Need: Word-frequency Masking Benefits Vision-Language Model Pre-training

2024-12-20 · Mingliang Liang, Martha Larson

Vision Language Models (VLMs) can be trained more efficiently if training sets can be reduced in size. Recent work has shown the benefits of masking text during VLM training using a variety of approaches: truncation, ran…

Language ModelingLanguage ModellingPOS

Masking Strategies for Image Manifolds

2016-06-15 · Hamid Dadkhahi, Marco F. Duarte

We consider the problem of selecting an optimal mask for an image manifold, i.e., choosing a subset of the pixels of the image that preserves the manifold's geometric structure present in the original data. Such masking …

Compressive Sensing

EntityBERT: Entity-centric Masking Strategy for Model Pretraining for the Clinical Domain

2021-06-01 · NAACL (BioNLP) 2021 6 · Chen Lin, Timothy Miller, Dmitriy Dligach, Steven Bethard 외

Transformer-based neural language models have led to breakthroughs for a variety of natural language processing (NLP) tasks. However, most models are pretrained on general domain data. We propose a methodology to produce…

NegationNegation DetectionRelationRelation Extraction+1

A hemispheric two-channel code accounts for binaural unmasking in humans

2021-11-08 · Jörg Encke, Mathias Dietz

Sound in noise is better detected or understood if target and masking sources originate from different locations. Mammalian physiology suggests that the neurocomputational process that underlies this binaural unmasking i…

Vocal Bursts Valence Prediction