Results 1 to 10 of about 6,103,025 (306)

Composition is the Core Driver of the Language-selective Network

Neurobiology of Language, 2020
The frontotemporal language network responds robustly and selectively to sentences. But the features of linguistic input that drive this response and the computations that these language areas support remain debated.
Francis Mollica, Matthew Siegelman, Evgeniia Diachek, Steven T. Piantadosi, Zachary Mineroff, Richard Futrell, Hope Kean, Peng Qian, Evelina Fedorenko   +8 more
doaj   +2 more sources

Lookback supports semi-parallel, just-in-time processing in second language written composition. [PDF]

PLoS ONE
PurposeCompetent written production results from semi-parallel processes that flow from content determination to finger movement. Information from upstream processes that is not used immediately is easily lost.
Evgeny Chukharev, Jens Roeser, Mark Torrance   +2 more
doaj   +3 more sources

Joint attention and exogenous attention allocation during mother-infant interaction at 12 months associate with 24-month vocabulary composition [PDF]

Frontiers in Psychology
IntroductionEarly attentional processes are inherently linked with early parent-infant interactions and play a critical role in shaping cognitive and linguistic development.
Elena Capelli, Serena Grumi, Luisa Vercellino, Livio Provenzi, Livio Provenzi   +4 more
doaj   +2 more sources

InternLM-XComposer: A Vision-Language Large Model for Advanced Text-image Comprehension and Composition [PDF]

arXiv.org, 2023
We propose InternLM-XComposer, a vision-language large model that enables advanced image-text comprehension and composition. The innovative nature of our model is highlighted by three appealing properties: 1) Interleaved Text-Image Composition: InternLM ...
Pan Zhang, Xiaoyi Wang, Yuhang Cao, Chao Xu, Linke Ouyang, Zhiyuan Zhao, Shuangrui Ding, Songyang Zhang, Haodong Duan, Hang Yan, Xinyu Zhang, Wei Li, Jingwen Li, Kai Chen, Conghui He, Xingcheng Zhang, Y. Qiao, Da Lin, Jiaqi Wang   +18 more
semanticscholar   +1 more source

How Abilities in Large Language Models are Affected by Supervised Fine-tuning Data Composition [PDF]

Annual Meeting of the Association for Computational Linguistics, 2023
Large language models (LLMs) with enormous pre-training tokens and parameters emerge diverse abilities, including math reasoning, code generation, and instruction following. These abilities are further enhanced by supervised fine-tuning (SFT).
Guanting Dong, Hongyi Yuan, Keming Lu, Chengpeng Li, Mingfeng Xue, Dayiheng Liu, Wei Wang, Zheng Yuan, Chang Zhou, Jingren Zhou   +9 more
semanticscholar   +1 more source

LoraHub: Efficient Cross-Task Generalization via Dynamic LoRA Composition [PDF]

arXiv.org, 2023
Low-rank adaptations (LoRA) are often employed to fine-tune large language models (LLMs) for new tasks. This paper investigates LoRA composability for cross-task generalization and introduces LoraHub, a simple framework devised for the purposive assembly
Chengsong Huang, Qian Liu, Bill Yuchen Lin, Tianyu Pang, Chao Du, Min Lin   +5 more
semanticscholar   +1 more source

XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models [PDF]

North American Chapter of the Association for Computational Linguistics, 2023
Without proper safeguards, large language models will readily follow malicious instructions and generate toxic content. This risk motivates safety efforts such as red-teaming and large-scale feedback learning, which aim to make models both helpful and ...
Paul Röttger, Hannah Rose Kirk, Bertie Vidgen, Giuseppe Attanasio, Federico Bianchi, Dirk Hovy   +5 more
semanticscholar   +1 more source

Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning [PDF]

International Conference on Learning Representations, 2023
The popularity of LLaMA (Touvron et al., 2023a;b) and other recently emerged moderate-sized large language models (LLMs) highlights the potential of building smaller yet powerful LLMs.
Mengzhou Xia, Tianyu Gao, Zhiyuan Zeng, Danqi Chen   +3 more
semanticscholar   +1 more source

Composable languages for bioinformatics: the NYoSh experiment [PDF]

PeerJ, 2014
Language WorkBenches (LWBs) are software engineering tools that help domain experts develop solutions to various classes of problems. Some of these tools focus on non-technical users and provide languages to help organize knowledge while other ...
Manuele Simi, Fabien Campagne
doaj   +2 more sources

Writing Composition Problem in Arabic Language Learning Among Arabic Language Education Students

Langkawi: Journal of The Association for Arabic and English, 2022
Learning foreign languages, especially Arabic, is still an obstacle for non-Arabic speakers. Apart from the problem of grammatical gaps, and phonological differences between source and target languages, writing composition proficiency is also a problem ...
Zulaeha Zulaeha
doaj   +1 more source