Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 6 appraisals
PloS one
Three-dimensional shape cues affect human and artificial recognition systems differently
Humans and neural networks use shape and texture information differently. While humans weigh shape heavily in their ultimate classification decision, neural networks are more biased towards texture cues. Many tests of shape vs. texture bias have focused on shape recognition from an object's external contour. However, shape information is also conveyed through internal contours, shading, and attached shadows, especially when an object is viewed from noncanonical perspectives. Using models from ShapeNet, we created datasets of 120,000 texture-substituted images of objects from many viewpoints with and without shading and attached shadows. We tested humans' and several neural networks' ability to classify these objects by both their shape and their texture. Humans were much better at classifying texture-substituted objects by their shape than any network, although these differences were greater when shape was defined only by the external contour than when 3D cues were included. Our findings suggest that networks' texture bias is reduced when 3D cues are included in images. We next tested whether the inclusion of 3D cues benefitted humans and neural networks more for images of objects viewed from canonical or noncanonical perspectives. Consistent with earlier research, we found that 3D cues primarily benefitted humans for noncanonical images. For neural networks, the greatest performance gains were for canonical images. These findings suggest fundamental differences in how humans and networks use shading and attached shadows for object recognition. We argue that humans use these cues to infer objects' 3D structures while neural networks use them as another surface-level cue like texture.
31 May 2026
Read appraisal →Philosophical transactions. Series A, Mathematical, physical, and engineering sciences
A sentence is worth a thousand pictures: can large language models understand hum4n L4ngu4ge and the W0rld behind W0rds?
The current generation of large language models (LLMs) has been linked to claims about human-like linguistic performance, and their applications are hailed both as a step towards artificial general intelligence and as a major advance in understanding the cognitive and even neural basis of human language. To assess these claims, first, we analysed the contribution of LLMs as theoretically informative representations of a target cognitive system versus atheoretical mechanistic tools. Second, we evaluated the models' ability to see the bigger picture through top-down feedback from higher levels of processing, which requires grounding in previous expectations and past world experience. We hypothesize that since models lack grounded cognition, they cannot take advantage of these features and instead solely rely on fixed associations between represented words and word vectors. To assess this, we ran a novel leet task (l33t t4sk), which requires decoding sentences in which letters are systematically replaced by numbers. In line with our hypothesis, the results suggest that humans excel in this task, whereas models struggle. We interpret these results by identifying the key abilities that are still missing from the current state of development of these models, which require solutions that go beyond increased system scaling. This article is part of the theme issue 'World models in natural and artificial intelligence'.
15 May 2026
Read appraisal →Philosophical transactions. Series A, Mathematical, physical, and engineering sciences
Empowerment gain and causal model construction: children and adults are sensitive to controllability and variability in their causal interventions
Learning about the causal structure of the world is a fundamental problem for human cognition. Causal models and especially causal learning have proved to be difficult for large pretrained models using standard techniques of deep learning. In contrast, cognitive scientists have applied advances in our formal understanding of causation in computer science, particularly within the causal Bayes net formalism, to understand human causal learning. In the very different tradition of reinforcement learning (RL), researchers have described an intrinsic reward signal called 'empowerment' which maximizes mutual information between actions and their outcomes. Empowerment may be an important bridge between classical Bayesian causal learning and RL and may help to characterize causal learning in humans and enable it in machines. If an agent learns an accurate causal world model, they will necessarily increase their empowerment, and increasing empowerment will lead to a more accurate causal world model. Empowerment may also explain distinctive features of children's causal learning, as well as providing a more tractable computational account of how that learning is possible. In an empirical study, we systematically test how children and adults use cues to empowerment to infer causal relations and design effective causal interventions. This article is part of the theme issue 'World models in natural and artificial intelligence'.
15 May 2026
Read appraisal →Philosophical transactions. Series A, Mathematical, physical, and engineering sciences
What physics offers for artificial intelligence? Lessons from the brain's inner time and its dynamics.
Computing can take advantage of physics-that is, 'whatever physics offers' (Jaeger et al. 2023 Nat. Commun. 14, 4911 (doi:10.1038/s41467-023-40533-1)). One of the key features physics provides is time: it describes rules and makes predictions about how activity patterns change over time, as governed by dynamics. We argue that computing can learn from physics in how it conceives time-namely, in terms of dynamics, i.e. the changing patterns of activity unfolding over time. In particular, we focus on the brain's intrinsic neural dynamics of spontaneous activity and how it 'uses' them for dynamic input processing and encoding in order to 'participate' in the world's physical time. By 'participate', we mean becoming part of the input dynamics: for instance, when listening or dancing to music, neural activity (via entrainment) and, consequently, mental activity follows the rhythm and dynamics of the music. This shapes the listener's experience, such as consciousness-the brain, and thus the subject, actively participates in music rhythms through temporal alignment (Northoff et al. 2023 Interface Focus13, 20220076 (doi:10.1098/rsfs.2022.0076)). Drawing on recent empirical evidence, we show that dynamic features such as the brain's scale-free activity and variability-which reflect an intrinsic temporal structure, or the brain's 'inner time'-are central to tracking and encoding input dynamics. Importantly, this processing is actively modulated rather than passively received, through the brain's own 'hidden dynamic repertoire'. Extending earlier discussions by Dreyfus and others, we argue that current computing devices, whether classical or natural (i.e. non-von Neumann machines), lack spontaneous activity and an inner time that can exert an active, rather than purely passive, influence on processing. As a result, they can neither actively process and encode input dynamics through their own inner time, nor 'use' or 'participate' in the dynamics of the world's physical time. Instead of 'being in time' and 'being in the world', current computing devices-and, by extension, artificial intelligence-are effectively 'locked out of time and world', meaning they are not directly connected to physical time. Unlike humans, they therefore cannot be characterized as 'being in time' or 'being in the world', which in turn prevents them from acquiring tacit or implicit knowledge, including the ability to navigate and behave flexibly within a continuously changing world. This article is part of the theme issue 'World models in natural and artificial intelligence'.
15 May 2026
Read appraisal →Proceedings of the National Academy of Sciences of the United States of America
Discovering regularity and mechanisms of word sense acquisition in childhood.
How does language use inform the emergence of word meanings in early life? Prior work in developmental psychology and the cognitive sciences typically focuses on studying word acquisition in children without specifying how different senses within a word emerge through time. To shed light on word sense acquisition, we propose a framework grounded in state-of-the-art computational methodologies of contextual word embedding to characterize how different senses of a word unfold as children acquire their lexicon. Our framework identifies word senses automatically by forming semantic clusters through natural language use, and analyzes 1,270 words from approximately 4 million utterances produced by children (19 to 144 mo) and their caretakers. The psychological validity of these senses was assessed based on a combination of dictionaries, human judgment of semantic similarity, and evaluation from a large language model. We then tested three hypotheses motivated by existing work on word sense emergence in language evolution. First, concrete senses of a word tend to emerge earlier than more abstract senses in child language. Second, word senses grow incrementally in semantic space across development. Third, algorithms of semantic chaining-how words spawn new senses by extending from existing senses-recapitulate the order of word senses in development. We find support for all three hypotheses. Our work suggests that the development of word meaning in language acquisition resembles that in language evolution and offers a converging view on the cognitive principles and mechanisms across timescales in the ontogeny and phylogeny of word sense emergence.
13 May 2026
Read appraisal →PloS one
Vision Transformer attention alignment with human visual perception in aesthetic object evaluation
Visual attention mechanisms play a crucial role in human perception and aesthetic evaluation. Recent advances in Vision Transformers (ViTs) have demonstrated remarkable capabilities in computer vision tasks, yet their alignment with human visual attention patterns remains underexplored, particularly in aesthetic contexts. This study investigates the correlation between human visual attention and ViT attention mechanisms when evaluating handcrafted objects. We conducted an eye-tracking experiment with 30 participants (9 female, 21 male, mean age 24.6 years) who viewed 20 artisanal objects comprising basketry bags and ginger jars. Using a Pupil Labs eye-tracker, we recorded gaze patterns and generated heatmaps representing human visual attention. Simultaneously, we analyzed the same objects using a pre-trained ViT model with DINO (Self-DIstillation with NO Labels), extracting attention maps from each of the 12 attention heads. We compared human and ViT attention distributions using four complementary metrics-Kullback-Leibler divergence, Structural Similarity Index (SSIM), Pearson's Correlation Coefficient (CC), and Similarity (SIM)-across varying Gaussian parameters ([Formula: see text]), yielding 1,152,000 distance evaluations. Additionally, we performed Areas of Interest (AOI) analysis to quantify ViT attention concentration within object regions. Statistical analysis revealed optimal correlation at [Formula: see text], with attention head #12 showing the strongest alignment with human visual patterns across all metrics. Significant differences were found between attention heads, with heads #7 and #9 demonstrating the greatest divergence from human attention ([Formula: see text]), Tukey HSD test). AOI analysis confirmed that all ViT heads concentrated attention significantly more within object regions than background areas ([Formula: see text]), with heads #12, #1, and #3 achieving lift values of +30 to +40 percentage points. Results indicate that while ViTs exhibit more global attention patterns compared to human focal attention, certain attention heads can approximate human visual behavior, particularly for specific object features like buckles in basketry items. These findings suggest potential applications of ViT attention mechanisms in product design and aesthetic evaluation, while highlighting fundamental differences in attention strategies between human perception and current AI models.
6 Apr 2026
Read appraisal →