Coco-lm: Correcting and contrasting text sequences for language model pretraining

Y Meng, C **ong, P Bajaj, P Bennett… - Advances in Neural …, 2021 - proceedings.neurips.cc
We present a self-supervised learning framework, COCO-LM, that pretrains Language
Models by COrrecting and COntrasting corrupted text sequences. Following ELECTRA-style …

PAQ: 65 million probably-asked questions and what you can do with them

P Lewis, Y Wu, L Liu, P Minervini, H Küttler… - Transactions of the …, 2021 - direct.mit.edu
Abstract Open-domain Question Answering models that directly leverage question-answer
(QA) pairs, such as closed-book QA (CBQA) models and QA-pair retrievers, show promise in …

Read before generate! faithful long form question answering with machine reading

D Su, X Li, J Zhang, L Shang, X Jiang, Q Liu… - arxiv preprint arxiv …, 2022 - arxiv.org
Long-form question answering (LFQA) aims to generate a paragraph-length answer for a
given question. While current work on LFQA using large pre-trained model for generation …

Salient span masking for temporal understanding

JR Cole, A Chaudhary, B Dhingra… - arxiv preprint arxiv …, 2023 - arxiv.org
Salient Span Masking (SSM) has shown itself to be an effective strategy to improve closed-
book question answering performance. SSM extends general masked language model …

Exploiting Abstract Meaning Representation for open-domain question answering

C Wang, Z Xu, Q Guo, X Hu, X Bai, Z Zhang… - arxiv preprint arxiv …, 2023 - arxiv.org
The Open-Domain Question Answering (ODQA) task involves retrieving and subsequently
generating answers from fine-grained relevant passages within a database. Current systems …

On the influence of masking policies in intermediate pre-training

Q Ye, BZ Li, S Wang, B Bolte, H Ma, W Yih… - arxiv preprint arxiv …, 2021 - arxiv.org
Current NLP models are predominantly trained through a two-stage" pre-train then fine-tune"
pipeline. Prior work has shown that inserting an intermediate pre-training stage, using …

Saaml: A framework for semi-supervised affective adaptation via metric learning

M Tran, Y Kim, CC Su, CH Kuo… - Proceedings of the 31st …, 2023 - dl.acm.org
Socially intelligent systems such as home robots should be able to perceive emotions and
social behaviors. Affect recognition datasets have limited labeled data, and existing large …

Investigating the Gap Between Single-Hop and Multi-Hop Questions in Closed-Book Question Answering via Question Decomposition

T Alkhaldi, C Chu, S Kurohashi - International Symposium on Distributed …, 2023 - Springer
Transformer-based language models (LMs) have been shown to perform question
answering (QA) competitively even when removing context and using only questions as …

[PDF][PDF] クローズドブック質問応答における言語モデルの知識強化

矢嶋梨穂 - 法政大学大学院紀要. 情報科学研究科編, 2024 - hosei.ecats-library.jp
Abstract Language models such as Chat-GPT generate new answers from the data they
learn, improving work efficiency and generating ideas. On the other hand, incorrect answers …

構造化知識を内包する自然言語理解システムの構築

矢嶋梨穂, 藤田悟 - 第 85 回全国大会講演論文集, 2023 - ipsj.ixsq.nii.ac.jp
論文抄録 ニューラルネットワークを用いた質問応答では, 機械読解やオープン検索質問応答などが
研究され, **年はクローズドブック質問応答の研究が進められている. 本研究では …