본문 바로가기
  • Machine Learning Paper Reviews (Mostly NLP)
STILTs: Supplementary Training on Pretrained Sentence Encoders Sentence Encoders on STILTs: Supplementary Training on Intermediate Labeled-data Tasks Jason Phang, Thibault F'evry, Samuel R. Bowman 27 Feb 2019 A Second Stage of Pretraining In order to overcome the flaws of existing encoders, transfer learning has allowed us to enhance the performance of various tasks by starting with a pretrained model that has already learned relevant features or representa.. 2023. 8. 26.
Predicting Spans Rather Than Tokens On BERT SpanBERT: Improving Pre-training by Representing and Predicting Spans Allen School of Computer Science & Engineering, University of Washington, Seattle, WA, etc. 18 Jan 2020 SpanBERT Coreference task is the task of finding all expressions that refer to the same entity in a text. For example, given a text as follows: "I voted for Nadar because he was most aligned with my values", she said. 'I', '.. 2023. 4. 29.
Representation Learning Basic (BERT) BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding Google AI Language 24 May 2019 A New Language Representation Model "A good representation is one that makes a subsequent learning task easier." The paper presents BERT(Bidirectional Encoder Representations from Transformer) which is designed to deeply learn the representations from unlabeled text on both left and ri.. 2023. 4. 25.
Morphological Capability of BERT DagoBERT: Generating Derivational Morphology with a Pretrained Language Model Valentin Hofmann, Janet B. Pierrehumbert, Hinrich schutze 7 Oct 2020 Derivational Morphology rather than Syntax and Semantics Among all those linguistic knowledges, syntax and semantics came into the lime light in NLP. The paper presents a study about the derivational morphological capability of BERT, suggesting a full.. 2023. 4. 16.
How Transformers Learn Long Sequences Taming Transformers for High-Resolution Image Synthesis Patrick Esser, Robin Rombach, Bjorn Ommer 23 Jun 2021 Transformer: Exploiting Its Highly Promising Learning Capabilities Since the paper has been published, many tasks relied on transformer architecture in various fields such as natural language processing or computer vision. However, because of its complex network which adapts complex rela.. 2023. 4. 5.
What Is Wrong With Backpropagation The Forward-Forward Algorithm: Some Preliminary Investigations Geoffrey Hinton [Google Brain] 27 Dec 2022 What Is Wrong With Backpropagation Despite the mathematical advantages we've obtained thanks to backpropagation, the paper maintains that backpropagation is an implausible method when we consider how the actual cortex is trained. Cortex does not mirror bottom-up connections like backpropagat.. 2023. 3. 25.