Results 11 to 20 of about 6,202,666 (342)
Lost in the Middle: How Language Models Use Long Contexts [PDF]
Transactions of the Association for Computational Linguistics, 2023 While recent language models have the ability to take long contexts as input, relatively little is known about how well they use longer context. We analyze the performance of language models on two tasks that require identifying relevant information in ...
Nelson F. Liu +6 more
semanticscholar +1 more source
Neural Video Compression with Diverse Contexts [PDF]
Computer Vision and Pattern Recognition, 2023 For any video codecs, the coding efficiency highly relies on whether the current signal to be encoded can find the relevant contexts from the previous reconstructed signals.
Jiahao Li, Bin Li, Yan Lu
semanticscholar +1 more source
Adapting Language Models to Compress Contexts [PDF]
Conference on Empirical Methods in Natural Language Processing, 2023 Transformer-based language models (LMs) are powerful and widely-applicable tools, but their usefulness is constrained by a finite context window and the expensive computational cost of processing long text documents.
Alexis Chevalier +3 more
semanticscholar +1 more source
TCTrack: Temporal Contexts for Aerial Tracking [PDF]
Computer Vision and Pattern Recognition, 2022 Temporal contexts among consecutive frames are far from being fully utilized in existing visual trackers. In this work, we present TCTrack11https://github.com/vision4robotics/TCTrack, a comprehensive framework to fully exploit temporal contexts for ...
Ziang Cao +5 more
semanticscholar +1 more source
OctAttention: Octree-based Large-scale Contexts Model for Point Cloud Compression [PDF]
AAAI Conference on Artificial Intelligence, 2022 In point cloud compression, sufficient contexts are significant for modeling the point cloud distribution. However, the contexts gathered by the previous voxel-based methods decrease when handling sparse point clouds.
Chunyang Fu +4 more
semanticscholar +1 more source
SituatedQA: Incorporating Extra-Linguistic Contexts into QA [PDF]
Conference on Empirical Methods in Natural Language Processing, 2021 Answers to the same question may change depending on the extra-linguistic contexts (when and where the question was asked). To study this challenge, we introduce SituatedQA, an open-retrieval QA dataset where systems must produce the correct answer to a ...
Michael J.Q. Zhang, Eunsol Choi
semanticscholar +1 more source
Adapting interventions to new contexts—the ADAPT guidance
British medical journal, 2021 Implementing interventions with a previous evidence base in new contexts might be more efficient than developing new interventions for each context. Although some interventions transfer well, effectiveness and implementation often depend on the context ...
G. Moore +15 more
semanticscholar +1 more source
Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contexts [PDF]
Computer Vision and Pattern Recognition, 2020 The rapid progress in 3D scene understanding has come with growing demand for data; however, collecting and an-notating 3D scenes (e.g. point clouds) are notoriously hard. For example, the number of scenes (e.g.
Ji Hou +3 more
semanticscholar +1 more source
Learning to Compose Dynamic Tree Structures for Visual Contexts [PDF]
Computer Vision and Pattern Recognition, 2018 We propose to compose dynamic tree structures that place the objects in an image into a visual context, helping visual reasoning tasks such as scene graph generation and visual Q&A.
Kaihua Tang +4 more
semanticscholar +1 more source
Frontiers in Psychology, 2022 In this manuscript, we introduce a theoretical model of climate radicalization that integrates social psychological theories of perceived unfairness with historical insights on radicalization to contribute to the knowledge of individuals’ processes of ...
Amarins Jansma +3 more
doaj +1 more source