On the influence of discourse connectives on the predictions of humans and language models DOI Creative Commons
James Britton, Yan Cong, Yu-Yin Hsu

и другие.

Frontiers in Human Neuroscience, Год журнала: 2024, Номер 18

Опубликована: Сен. 30, 2024

Psycholinguistic literature has consistently shown that humans rely on a rich and organized understanding of event knowledge to predict the forthcoming linguistic input during online sentence comprehension. We, authors, expect sentences maintain coherence with preceding context, making congruent sequences easier process than incongruent ones. It is widely known discourse relations between (e.g., temporal, contingency, comparison) are generally made explicit through specific particles, as connectives , and, but, because, after ). However, some easily accessible speakers, given their knowledge, can also be left implicit. The goal this paper investigate importance in prediction events human language processing pretrained models, focus concessives contrastives, which signal comprehenders event-related predictions have reversed . Inspired by previous work, we built comprehensive set story stimuli Italian Mandarin Chinese differ plausibility situation being described presence or absence connective. We collected judgments reading times from native speakers for stimuli. Moreover, correlated results experiments computational modeling, using Surprisal scores obtained via Transformer-based models. judgements were seven-point Likert scale analyzed cumulative link mixed modeling (CLMM), while model surprisal linear effects regression (LMER). found NLMs sensitive connectives, although they struggle reproduce expectation reversal due connective changing scenario; even less aligned data, no either Surprisal.

Язык: Английский

The Breakthrough of Large Language Models Release for Medical Applications: 1-Year Timeline and Perspectives DOI Creative Commons
Marco Cascella, Federico Semeraro, Jonathan Montomoli

и другие.

Journal of Medical Systems, Год журнала: 2024, Номер 48(1)

Опубликована: Фев. 17, 2024

Within the domain of Natural Language Processing (NLP), Large Models (LLMs) represent sophisticated models engineered to comprehend, generate, and manipulate text resembling human language on an extensive scale. They are transformer-based deep learning architectures, obtained through scaling model size, pretraining corpora, computational resources. The potential healthcare applications these primarily involve chatbots interaction systems for clinical documentation management, medical literature summarization (Biomedical NLP). challenge in this field lies research diagnostic decision support, as well patient triage. Therefore, LLMs can be used multiple tasks within care, research, education. Throughout 2023, there has been escalation release LLMs, some which applicable domain. This remarkable output is largely effect customization pre-trained like chatbots, virtual assistants, or any system requiring human-like conversational engagement. As professionals, we recognize imperative stay at forefront knowledge. However, keeping abreast rapid evolution technology practically unattainable, and, above all, understanding its limitations remains a subject ongoing debate. Consequently, article aims provide succinct overview recently released emphasizing their use medicine. Perspectives more range safe effective also discussed. upcoming evolutionary leap involves transition from AI-powered designed answering questions versatile practical tool providers such generalist biomedical AI multimodal-based calibrated decision-making processes. On other hand, development accurate partners could enhance engagement, offering personalized improving chronic disease management.

Язык: Английский

Процитировано

71
Driving and suppressing the human language network using large language models DOI
Greta Tuckute, Aalok Sathe, Shashank Srikant

и другие.

Nature Human Behaviour, Год журнала: 2024, Номер 8(3), С. 544 - 561

Опубликована: Янв. 3, 2024

Язык: Английский

Процитировано

30
Language Model Behavior: A Comprehensive Survey DOI Creative Commons

Tyler A. Chang,

Benjamin Bergen

Computational Linguistics, Год журнала: 2023, Номер 50(1), С. 293 - 350

Опубликована: Ноя. 15, 2023

Abstract Transformer language models have received widespread public attention, yet their generated text is often surprising even to NLP researchers. In this survey, we discuss over 250 recent studies of English model behavior before task-specific fine-tuning. Language possess basic capabilities in syntax, semantics, pragmatics, world knowledge, and reasoning, but these are sensitive specific inputs surface features. Despite dramatic increases quality as scale hundreds billions parameters, the still prone unfactual responses, commonsense errors, memorized text, social biases. Many weaknesses can be framed over-generalizations or under-generalizations learned patterns text. We synthesize results highlight what currently known about large capabilities, thus providing a resource for applied work research adjacent fields that use models.

Язык: Английский

Процитировано

39
Large language models in psychiatry: Opportunities and challenges DOI
Sebastian Volkmer, Andreas Meyer‐Lindenberg, Emanuel Schwarz

и другие.

Psychiatry Research, Год журнала: 2024, Номер 339, С. 116026 - 116026

Опубликована: Июнь 12, 2024

Язык: Английский

Процитировано

12
Driving and suppressing the human language network using large language models DOI Creative Commons
Greta Tuckute, Aalok Sathe, Shashank Srikant

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Апрель 16, 2023

Transformer models such as GPT generate human-like language and are highly predictive of human brain responses to language. Here, using fMRI-measured 1,000 diverse sentences, we first show that a GPT-based encoding model can predict the magnitude response associated with each sentence. Then, use identify new sentences predicted drive or suppress in network. We these model-selected novel indeed strongly activity areas individuals. A systematic analysis reveals surprisal well-formedness linguistic input key determinants strength These results establish ability neural network not only mimic but also noninvasively control higher-level cortical areas, like

Язык: Английский

Процитировано

14
Evaluating ChatGPT’s Consciousness and Its Capability to Pass the Turing Test: A Comprehensive Analysis DOI Open Access
Matjaž Gams,

Sebastjan Kramar

Journal of Computer and Communications, Год журнала: 2024, Номер 12(03), С. 219 - 237

Опубликована: Янв. 1, 2024

This study explores the capabilities of ChatGPT, specifically in relation to consciousness and its performance Turing Test. The article begins by examining diverse perspectives among both cognitive AI researchers regarding ChatGPT's ability pass It introduces a hierarchical categorization test versions, suggesting that ChatGPT approaches success test, albeit primarily with na?ve users. Expert users, conversely, can easily identify limitations. paper presents various theories consciousness, particular focus on Integrated Information Theory proposed Tononi. theory serves as framework for assessing level consciousness. Through an evaluation based five axioms theorems IIT, finds surpasses previous systems certain aspects; however, significantly falls short achieving particularly when compared biological sentient beings. concludes emphasizing importance recognizing similar generative models highly advanced intelligent tools, yet distinctly lacking attributes found living organisms.

Язык: Английский

Процитировано

5
A Better Way to Do Masked Language Model Scoring DOI Creative Commons
Carina Kauf, А. Е. Иванова

Опубликована: Янв. 1, 2023

Estimating the log-likelihood of a given sentence under an autoregressive language model is straightforward: one can simply apply chain rule and sum values for each successive token. However, masked models (MLMs), there no direct way to estimate sentence. To address this issue, Salazar et al. (2020) propose pseudo-log-likelihood (PLL) scores, computed by successively masking token, retrieving its score using rest as context, summing resulting values. Here, we demonstrate that original PLL method yields inflated scores out-of-vocabulary words adapted metric, in which mask not only target but also all within-word tokens right target. We show our metric (PLL-word-l2r) outperforms both are masked. In particular, it better satisfies theoretical desiderata correlates with from models. Finally, choice affects even tightly controlled, minimal pair evaluation benchmarks (such BLiMP), underscoring importance selecting appropriate scoring evaluating MLM properties.

Язык: Английский

Процитировано

11
An LLM-Based Inventory Construction Framework of Urban Ground Collapse Events with Spatiotemporal Locations DOI Creative Commons
Yanan Hao, Qi Jin, Xiaowen Ma

и другие.

ISPRS International Journal of Geo-Information, Год журнала: 2024, Номер 13(4), С. 133 - 133

Опубликована: Апрель 16, 2024

Historical news media reports serve as a vital data source for understanding the risk of urban ground collapse (UGC) events. At present, application large language models (LLMs) offers unprecedented opportunities to effectively extract UGC events and their spatiotemporal information from vast amount data. Therefore, this study proposes an LLM-based inventory construction framework consisting three steps: crawling, event recognition, attribute extraction. Focusing on Zhejiang province, China, test region, total 27 cases 637 were collected 11 prefecture-level cities. The method achieved recall rate over 60% precision below 35%, indicating its potential automatically screening events; however, accuracy needs be improved account confusion with other events, such bridge collapses. obtained is first open access based internet reports, dates locations, co-ordinates derived unstructured contents. Furthermore, provides insights into spatial pattern frequency in supplementing statistical provided by local government.

Язык: Английский

Процитировано

4
A survey on knowledge-enhanced multimodal learning DOI Creative Commons
Maria Lymperaiou, Giorgos Stamou

Artificial Intelligence Review, Год журнала: 2024, Номер 57(10)

Опубликована: Сен. 9, 2024

Язык: Английский

Процитировано

4
Shades of zero: Distinguishing impossibility from inconceivability DOI

Jennifer Hu,

Felix Sosa,

Tomer Ullman

и другие.

Journal of Memory and Language, Год журнала: 2025, Номер 143, С. 104640 - 104640

Опубликована: Апрель 15, 2025

Язык: Английский

Процитировано

0