Prior-data gravity and framework reabsorption: a behavioral theory of novelty displacement in large language models

An LLM, or large language model, is a neural network trained on vast text to predict the next token, the engine behind ChatGPT, Gemini, Claude and AI search.

An LLM, or large language model, is a neural network trained on vast text to predict the next token, the engine behind ChatGPT, Gemini, Claude and AI search.

Background: Recent advancements in digital health technologies have led to the accumulation of large volumes of unstructured text data, particularly in electronic health records. Given that #nursing documentation frequently includes detailed descriptions of daily living activities and psychosocial contexts, existing medical natural language processing (NLP) tools may not be directly applicable without adaptation. Moreover, the integration of standardized #nursing terminologies into NLP systems remains limited. Although NLP has the potential to support risk detection, information management, and care planning, its practical applications in #nursing have not yet been systematically mapped. Objective: This scoping review aimed to map the practical applications of NLP in #nursing practice, with a specific focus on real-world clinical use, particularly for assessment, documentation, and interventions, and to identify the study characteristics, application domains, outcomes, and challenges. Methods: Following the Joanna Briggs Institute methodology and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines, a systematic search was conducted across PubMed, CINAHL, the Cochrane Library, Scopus, Embase, and Ichushi-Web in October 2023 and June 2025. The population-concept-context framework (population: #nurses; concept: NLP applications; context: clinical settings) was used to guide study selection. Two independent reviewers screened 2889 unique records and reviewed the full texts of 72 studies. Data were descriptively synthesized and thematically classified. Results: Six studies met the inclusion criteria, indicating that empirically evaluated applications of NLP in real-world #nursing practice remain limited. NLP applications were categorized into three domains: (1) text summarization and generation, (2) information extraction and classification, and (3) real-time detection and screening. NLP was applied across multiple stages of #nursing practice, including handovers, assessment, care planning, continuity of care, and patient safety management. Notably, NLP was used as a form of augmented intelligence to complement, rather than replace, #nurses’ clinical judgment and to support clinical decision-making. These findings also highlight a substantial gap between rapid technological advances and the limited implementation of NLP in real-world #nursing practice. Conclusions: This review highlights the substantial gap between rapid technological advances and the limited implementation of NLP in real-world #nursing practice. NLP may help reduce manual review workload, support early risk identification, and facilitate individualized care planning in selected #nursing contexts. However, direct empirical evidence regarding its effects on #nursing workflows, #nurses’ clinical decision-making, and patient-level outcomes remains limited. Future research should prioritize real-world implementation studies and evaluate #nurses’ acceptance, adoption barriers, workflow integration, and the influence of NLP on clinical decision-making.

Sonnet 5.5 cache reads halve to $0.10 per million, and Team plans gain up to $500 a month in API credits. What changes for teams running agents at high volume?

Sonnet 5.5 cache reads halve to $0.10 per million, and Team plans gain up to $500 a month in API credits. What changes for teams running agents at high volume?

Background: Subjective cognitive decline (SCD) typically refers to self- or informant-reported decline in cognition despite the absence of objective impairment on standardized testing. Older adults with documented normal cognitive test performance provide a pragmatic anchor cohort for electronic health record (EHR)–based SCD phenotyping. However, cognitive concerns are primarily recorded in unstructured notes and are inconsistently documented, making it unclear how often and in whom concerns are captured in routine care. Objective: This study aims to operationalize EHR-based SCD phenotyping in an objectively normal-testing cohort using a large language model (LLM)–based natural language processing approach and to examine the frequency and clinical and sociodemographic correlates of cognitive concern documentation. Methods: We conducted an EHR-based observational study of patients aged 65 years or older with a first normal cognitive test recorded in EHR flowsheets between January 2019 and April 2024 in a large health care system. We developed and iteratively refined a 2-stage LLM-based natural language processing pipeline to identify documented cognitive concerns in unstructured notes during the 12 months prior to the index date, and evaluated performance against manual review. We quantified the frequency of documented concerns within this normal-testing cohort and used multivariable logistic regression to assess associations with sociodemographic factors (age, marital status, insurance, and neighborhood Area Deprivation Index), sequentially adjusting for clinical comorbidities and relevant medications. Results: Among 15,750 older adults with normal cognitive test scores, 13.8% (n=2175) had at least 1 documented cognitive concern in the prior year, captured in 1.2% (7394/605,177) of notes. On manual validation, the 2-stage pipeline (Med42-v2-8B screening followed by GPT-4o confirmation) achieved a sensitivity of 0.957, positive predictive value of 0.935, specificity of 0.985, and -score of 0.945 for cognitive concern identification. Documentation was more likely in older individuals, those with commercial (vs Medicare) insurance, and those with neurological and psychiatric conditions. In fully adjusted models, Parkinson disease (adjusted odds ratio [aOR] 5.29, 95% CI 3.60‐7.77), traumatic brain injury (aOR 4.63, 95% CI 3.15‐6.81), stroke or transient ischemic attack (aOR 3.46, 95% CI 3.15‐4.18), epilepsy (aOR 2.52, 95% CI 1.83‐3.49), depression (aOR 1.54, 95% CI 1.36‐1.75), and excessive alcohol use (aOR 1.48, 95% CI 1.06‐2.06) were among the strongest correlates of cognitive concern documentation. In contrast, obesity (aOR 0.73, 95% CI 0.65‐0.82), hyperlipidemia (aOR 0.59, 95% CI 0.51‐0.67), and residence in more deprived neighborhoods (higher Area Deprivation Index) were associated with lower odds of documented cognitive concerns. Conclusions: A 2-stage LLM pipeline enabled accurate identification of documented cognitive concerns consistent with SCD among older adults with normal cognitive testing. Documentation was uncommon and selectively captured by clinical and sociodemographic factors, with implications for equity and the validity of EHR-based phenotypes.

Retrieval-augmented generation (RAG) makes a language model fetch documents before it answers. How the pipeline works, where it fails and why publishers sue.