Oldest Papers

Date: Oldest Clear all

Advanced filtersTopic: All

Experimental

Topic

Topics are auto-detected from title, abstract, and metadata and may be imperfect.

Publication Date

Newest Oldest Clear

Average Rating

Highest Lowest

<scp> CO <sub>2</sub> </scp> Rise Directly Impairs Crop Nutritional Quality

Sterre F. ter Haar, Peter M. van Bodegom, Laura Scherer

Global Change Biology·2025

ABSTRACT Rising CO 2 levels indirectly affect food availability (e.g., through climate shifts, biological consequences, and shifting socioeconomic systems). Less known is the direct effect of CO 2 on nutrient availability by changing the plant stoichiometry. We help fill this gap, presenting a new methodology and creating the largest meta‐analysis on edible plant parts to date, including the most data (5324 entries covering 29,524 observation pairs), crops (43), and (anti)nutrients (32). The results show a pervasive elemental shift across a wide range of species as CO 2 rises. Elements respond differentially, with zinc decreasing the most. Plants with not only a C3 but also a C4 photosynthetic pathway respond. Impacts vary by species and tissue type. This stoichiometric shift can lead to (worsening) malnutrition even in previously sufficient populations. Nutrient security is under threat even if food security remains adequate; food will become more caloric and less nutritious.

No ratings yet

View paper →

Integration of fisheries and ecological data to support spatial management: The case of blackspot seabream (Pagellus bogaraveo) in the western Mediterranean Sea

Diego Panzeri, Juan Gil Herrera, Cristina García-Ruíz, Lucia Rueda, Meryem Benziane, Mohammed Idrissi Malouli, Pilar Hernández, Simone Libralato

Fisheries Research·2025

No abstract available

No ratings yet

View paper →

Inferring the sensitivity of wastewater metagenomic sequencing for early detection of viruses: a statistical modelling study

Simon L Grimm, Jeff T Kaufman, Daniel P Rice, Charles Whittaker, William J Bradshaw, Michael R McLaren

The Lancet Microbe·2025·1 citations

No abstract available

No ratings yet

View paper →

The hidden biodiversity knowledge split in biological collections

Gabriel Nakamura, Bruno Henrique Mioto Stabile, Livia Estéfane Fernandes Frateles, Matheus Lima Araujo, Emanuel Bruno Neuhaus, Manoela Maria Ferreira Marinho, Melina de Souza Leite, Aline Richter, Liuyong Ding, Tiago Magalhães da Silva Freitas, Bruno Soares, Weferson Júnio da Graça, Mario R. Moura, José Alexandre Felizola Diniz-Filho

Proceedings of the Royal Society B: Biological Sciences·2025·1 citations

Ecological and evolutionary processes generate biodiversity, yet how biodiversity data are organized and shared globally can shape our understanding of these processes. We show that name-bearing type specimens—the primary reference for species identity—of all freshwater and brackish fish species are predominantly housed in Global North museums, disconnected from their countries of origin. This geographical divide creates a ‘knowledge split’ with consequences for biodiversity science, particularly in the Global South, where researchers face barriers in studying native species’ name bearers housed abroad. Meanwhile, Global North collections remain flooded with non-native name bearers. We relate this imbalance to historical and socioeconomic factors, which ultimately restrict access to critical taxonomic reference materials and hinder global species documentation. To address this disparity, we call for international initiatives to promote fairer access to biological knowledge, including specimen repatriation, improved accessibility protocols for researchers in countries where specimens originated and inclusive research partnerships.

★ 5.0 (1)

View paper →

PreviousPage 374 of 507Next

Worse for Women, Bad for All: A 62-Nation Study Confirms and Extends Ambivalent Sexism Principles to Reveal Greater Social Dysfunction in Sexist Nations

Magdalena Zawisza, Natasza Kosakowska-Berezecka, Peter Glick, Michal Olech, Tomasz Besta, Paweł Jurek, Jurand Sobiecki, Deborah L. Best, Jennifer K. Bosson, Joseph A. Vandello, Saba Safdar, Anna Włodarczyk, Magdalena Żadkowska

Social Psychological and Personality Science·2025·10 citations

We retested core ambivalent sexism theory tenets and explored novel correlations with national outcomes in 62 nations. Replicating Glick et al., cross-national analyses supported (a) hostile sexism (HS) and benevolent sexism (BS) as cross-culturally recognizable, complementary ideologies associated with gender inequality; (b) women appearing to be influenced by, but also resisting men’s HS and embracing BS to counter men’s HS (outscoring men in some highly sexist nations). Novel cross-national comparisons showed (a) men’s HS and both genders’ BS correlated with fewer women in paid work, whereas only BS correlated with domestic labor inequity, (b) both HS and BS correlated with accepting intimate partner violence toward women. Finally, HS and BS correlated with generally dysfunctional national outcomes: antidemocratic tendencies, less productivity, more collective violence, and lower healthy lifespan for both genders. Results reinforce that BS harms women and suggest men also have a stake in reducing sexist ideologies.

No ratings yet

View paper →

Abolition: Is this the only pathway to upholding human rights and ensuring epistemic justice in psychiatry? A key informant qualitative study

Cath Roper, Nina Joffee-Kohn, Vrinda Edan, Natasha Swingler, Piers Gooding, Bridget Hamilton

International Journal of Law and Psychiatry·2025

Introduction: Mental health legislation authorises involuntary psychiatric intervention in certain circumstances. Although human rights concerns are becoming more prominent, debates among legal experts, clinicians and activists continue to swirl around people's rights to equal recognition before the law, such as described in the (United Nations Convention on the Rights of Persons with disabilities, 2006). Aim: This qualitative descriptive study aimed to better understand diverse views of people known to hold a critique of coercion in mental health services, on the practical expression of upholding human rights in the context of mental health laws. Method: Individual semi-structured interviews were conducted with 15 key informants from five different countries and data was analysed using an inductive, thematic approach. Results: Overall, informants characterised mental health laws as discriminatory, harmful and unjustifiable. Three themes and six sub-themes were identified. This study reports on the major themes which include: an ethical position (focusing on the present harms associated with mental health laws), strategies, (an expression of the opportunity to bring about change) and a visionary position. We explore these three features in the views of key informants as important positions in the field of abolition, and analyse each for the ‘hermeneutic resources’ – forms of collective interpretive resources - they provide. Conclusions: Abolition of mental health laws is often seen as not feasible in the context of psychiatry. However, abolition theories and practices are hermeneutic resources that need to be better understood because they offer social justice and community-led solutions beyond mental health laws and systems.

No ratings yet

View paper →

Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering

Eric Bigelow, Daniel Wurgaft, YingQiao Wang, Noah Goodman, Tomer Ullman, Hidenori Tanaka, Ekdeep Singh Lubana

arXiv·2025

Large language models (LLMs) can be controlled at inference time through prompts (in-context learning) and internal activations (activation steering). Different accounts have been proposed to explain these methods, yet their common goal of controlling model behavior raises the question of whether these seemingly disparate methodologies can be seen as specific instances of a broader framework. Motivated by this, we develop a unifying, predictive account of LLM control from a Bayesian perspective. Specifically, we posit that both context- and activation-based interventions impact model behavior by altering its belief in latent concepts: steering operates by changing concept priors, while in-context learning leads to an accumulation of evidence. This results in a closed-form Bayesian model that is highly predictive of LLM behavior across context- and activation-based interventions in a set of domains inspired by prior work on many-shot in-context learning. This model helps us explain prior empirical phenomena - e.g., sigmoidal learning curves as in-context evidence accumulates - while predicting novel ones - e.g., additivity of both interventions in log-belief space, which results in distinct phases such that sudden and dramatic behavioral shifts can be induced by slightly changing intervention controls. Taken together, this work offers a unified account of prompt-based and activation-based control of LLM behavior, and a methodology for empirically predicting the effects of these interventions.

No ratings yet

View paper →

Isotropic Curvature Model for Understanding Deep Learning Optimization: Is Gradient Orthogonalization Optimal?

Weijie Su

arXiv·2025

In this paper, we introduce a model for analyzing deep learning optimization over a single iteration by leveraging the matrix structure of the weights. We derive the model by assuming isotropy of curvature, including the second-order Hessian and higher-order terms, of the loss function across all perturbation directions; hence, we call it the isotropic curvature model. This model is a convex optimization program amenable to analysis, which allows us to understand how an update on the weights in the form of a matrix relates to the change in the total loss function. As an application, we use the isotropic curvature model to analyze the recently introduced Muon optimizer and other matrix-gradient methods for training language models. First, we show that under a general growth condition on the curvature, the optimal update matrix is obtained by making the spectrum of the original gradient matrix more homogeneous -- that is, making its singular values closer in ratio -- which in particular improves the conditioning of the update matrix. Next, we show that the orthogonalized gradient becomes optimal for the isotropic curvature model when the curvature exhibits a phase transition in growth. Taken together, these results suggest that the gradient orthogonalization employed in Muon and other related methods is directionally correct but may not be strictly optimal. Finally, we discuss future research on how to leverage the isotropic curvature model for designing new optimization methods for training deep learning and language models.

No ratings yet

View paper →

Cross-Validated Causal Inference: a Modern Method to Combine Experimental and Observational Data

Xuelin Yang, Licong Lin, Susan Athey, Michael I. Jordan, Guido W. Imbens

arXiv·2025

We develop new methods to integrate experimental and observational data in causal inference. While randomized controlled trials offer strong internal validity, they are often costly and therefore limited in sample size. Observational data, though cheaper and often with larger sample sizes, are prone to biases due to unmeasured confounders. To harness their complementary strengths, we propose a systematic framework that formulates causal estimation as an empirical risk minimization (ERM) problem. A full model containing the causal parameter is obtained by minimizing a weighted combination of experimental and observational losses--capturing the causal parameter's validity and the full model's fit, respectively. The weight is chosen through cross-validation on the causal parameter across experimental folds. Our experiments on real and synthetic data show the efficacy and reliability of our method. We also provide theoretical non-asymptotic error bounds.

No ratings yet

View paper →

Oldest Papers

<scp> CO <sub>2</sub> </scp> Rise Directly Impairs Crop Nutritional Quality

Integration of fisheries and ecological data to support spatial management: The case of blackspot seabream (Pagellus bogaraveo) in the western Mediterranean Sea

Inferring the sensitivity of wastewater metagenomic sequencing for early detection of viruses: a statistical modelling study

The hidden biodiversity knowledge split in biological collections

Worse for Women, Bad for All: A 62-Nation Study Confirms and Extends Ambivalent Sexism Principles to Reveal Greater Social Dysfunction in Sexist Nations

Abolition: Is this the only pathway to upholding human rights and ensuring epistemic justice in psychiatry? A key informant qualitative study

The Goon Squad, by Daniel Kolitz

Naive Reviews: John Donne’s Holy Sonnets, Part 1

Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering

Isotropic Curvature Model for Understanding Deep Learning Optimization: Is Gradient Orthogonalization Optimal?

Cross-Validated Causal Inference: a Modern Method to Combine Experimental and Observational Data

Heated Rivalry is anti-Dystopia art