Creativity is increasingly recognized as a core competency for the 21st century, making its development a priority in education, research, and industry. To effectively cultivate creativity, researchers and educators need reliable and accessible assessment tools. Recent software developments have significantly enhanced the administration and scoring of creativity measures; however, existing software often requires expertise in experiment design and computer programming, limiting its accessibility to many educators and researchers. In the current work, we introduce CAP-the Creativity Assessment Platform-a free web application for building creativity assessments, collecting data, and automatically scoring responses (cap.ist.psu.edu). CAP allows users to create custom creativity assessments in ten languages using a simple, point-and-click interface, selecting from tasks such as the Short Story Task, Drawing Task, and Scientific Creative Thinking Test. Users can automatically score task responses using machine learning models trained to match human creativity ratings-with multilingual capabilities, including the new Cross-Lingual Alternate Uses Scoring (CLAUS), a large language model achieving strong prediction of human creativity ratings in ten languages. CAP also provides a centralized dashboard to monitor data collection, score assessments, and automatically generate text for a Methods section based on the study's tasks, metrics, and instructions-with a single click-promoting transparency and reproducibility in creativity assessment. Designed for ease of use, CAP aims to democratize creativity measurement for researchers, educators, and everyone in between.

Mete Ismayilzada, Antonio Laverghetta Jr., Simone A. Luchini, Reet Patel, Antoine Bosselut, Lonneke Van Der Plas, Roger E. Beaty. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.

Creativity is hypothesized to arise from a mental state which balances spontaneous thought and cognitive control, corresponding to functional connectivity between the brain's Default Mode (DMN) and Executive Control (ECN) Networks. Here, we conduct a large-scale, multi-center examination of this hypothesis. Employing a meta-analytic network neuroscience approach, we analyze resting-state fMRI and creative task performance across 10 independent samples from Austria, Canada, China, Japan, and the United States (N = 2433)-constituting the largest and most ethnically diverse creativity neuroscience study to date. Using time-resolved network analysis, we investigate the relationship between creativity (i.e., divergent thinking ability) and dynamic switching between DMN and ECN. We find that creativity, but not general intelligence, can be reliably predicted by the number of DMN-ECN switches. Importantly, we identify an inverted-U relationship between creativity and the degree of balance between DMN-ECN switching, suggesting that optimal creative performance requires balanced brain network dynamics. Furthermore, an independent task-fMRI validation study (N = 31) demonstrates higher DMN-ECN switching during creative idea generation (compared to a control condition) and replicates the inverted-U relationship. Therefore, we provide robust evidence across multi-center datasets that creativity is tied to the capacity to dynamically switch between brain networks supporting spontaneous and controlled cognition.

Creativity assessment in science and engineering is increasingly based on both human and AI judgment, but the cognitive processes and biases behind these evaluations remain poorly understood. We conducted two experiments examining how including example solutions with ratings impact creativity evaluation, using a finegrained annotation protocol where raters were tasked with explaining their originality scores and rating for the facets of remoteness (whether the response is "far" from everyday ideas), uncommonness (whether the response is rare), and cleverness. In Study 1, we analyzed creativity ratings from 72 experts with formal science or engineering training, comparing those who received example solutions with ratings (example) to those who did not (no example). Computational text analysis revealed that, compared to experts with examples, no-example experts used more comparative language (e.g., "better/worse") and emphasized solution uncommonness, suggesting they may have relied more on memory retrieval for comparisons. In Study 2, parallel analyses with state-of-the-art LLMs revealed that models prioritized uncommonness and remoteness of ideas when rating originality, suggesting an evaluative process rooted around the semantic similarity of ideas. In the example condition, while LLM accuracy in predicting the true originality scores improved, the correlations of remoteness, uncommonness, and cleverness with originality also increased substantially -- to upwards of $0.99$ -- suggesting a homogenization in the LLMs evaluation of the individual facets. These findings highlight important implications for how humans and AI reason about creativity and suggest diverging preferences for what different populations prioritize when rating.

Complex cognition, such as creativity, relies on cognitive integration of various component processes (e.g., memory, attention, and imagery). Yet, current methods cannot fully capture how the brain integrates cognitive processes during complex tasks. Previous research suggests that communication between functionally dissimilar regions might underlie cognitive integration, allowing for complex cognition. Here, we provide a formal test of this notion using task-based fMRI (n = 28) to assess functional connectivity (FC) among sets of regions ("levels") varying in their functional dissimilarity (defined by differences in resting-state FC profiles) across five tasks hypothesized to vary in cognitive complexity. Each task involved conceptual association and/or idea generation. We found that as task complexity increased, task-FC between regions with greater functional dissimilarity also increased, and the strength of this linear trend positively predicted the relative complexity of tasks. Thus, more complex tasks recruited greater interactions between functionally dissimilar regions. Furthermore, this effect was primarily driven by the default mode and frontoparietal control networks, especially connector hubs within these networks. Task-FC at the highest functional dissimilarity levels was most related to metaphor production and bi-association (involving integrating two concepts), followed by generating novel object uses and uncommon association (involving expanding one concept), and was least related to common association (thus, this task was the least complex). Altogether, task-FC across functional dissimilarity levels robustly tracked the cognitive complexity of tasks, supporting the validity of this neural feature for measuring cognitive complexity in a continuous manner and for data-driven tests of theorized differences in task complexity.

Creativity is fundamentally a collaborative process. Yet as generative AI becomes increasingly integrated into creative work, understanding how AI reshapes collaboration has become critical. This pre-registered study directly compares human-human and human-AI collaboration dynamics across two creative tasks: the Alternative Uses Task (AUT) and creative short story writing. Participants were randomly assigned to pairs in either human-human (N = 68 pairs) or human-AI (GPT-4o; N = 72 pairs) conditions, with partners alternating turns as first responders to examine how initiation order shapes the creative process over time. Our findings reveal that the apparent "AI advantage" in creative collaboration is illusory, driven primarily by increased AI verbosity rather than enhanced creativity. Critically, collaboration with AI partners negatively impacted humans' own creative responses compared to human-human partnerships, with human-AI collaboration failing to enhance idea originality or diversity relative to human-human collaboration. Human partners demonstrated superior collaborative effectiveness that strengthened over time, indicating that current generative AI systems, while producing more verbose outputs, do not replicate the collective creativity characteristic of human-human collaboration. These results challenge assumptions about AI's creative potential, with direct implications for AI system design and collaborative creative practice.

Creativity is a key 21st-century skill and a consistent predictor of academic learning outcomes. Despite decades of research on creativity and learning, little is known about the cognitive mechanisms underlying their relationship. In two studies, we examined whether creativity supports associative learning through associative thinking-the ability to generate novel word associations-an ability central to creativity which has not been previously tied to associative learning. In Study 1, we found that students who generated more novel word associations learned more words on a foreign language learning test 24 h later. In Study 2, we replicated and extended the effect to naturalistic creativity tasks (i.e., writing short stories and sketching line drawings), finding associative thinking mediated the relationship between creativity and associative learning. Importantly, both studies controlled for general intelligence. Our findings suggest that creativity's contribution to learning operates partly through a shared cognitive capacity for making new connections.

Linking neurobiology to relatively stable individual differences in cognition, emotion, motivation, and behavior can require large sample sizes to yield replicable results. Given the nature of between-person research, sample sizes at least in the hundreds are likely to be necessary in most neuroimaging studies of individual differences, regardless of whether they are investigating the whole brain or more focal hypotheses. However, the appropriate sample size depends on the expected effect size. Therefore, we propose four strategies to increase effect sizes in neuroimaging research, which may help to enable the detection of replicable between-person effects in samples in the hundreds rather than the thousands: (1) theoretical matching between neuroimaging tasks and behavioral constructs of interest; (2) increasing the reliability of both neural and psychological measurement; (3) individualization of measures for each participant; and (4) using multivariate approaches with cross-validation instead of univariate approaches. We discuss challenges associated with these methods and highlight strategies for improvements that will help the field to move toward a more robust and accessible neuroscience of individual differences.

Creativity is a fundamental ability of humans, vital to most domains of work and art. With the rising popularity of generative AI, research is needed to determine its costs and benefits for human creativity. We conducted a naturalistic study with university students (N = 263) who wrote short-stories with and without the help of ChatGPT. Stories written with AI assistance were significantly more original, but also more semantically similar to each other (i.e., less diverse). Critically, using a novel analysis of participant-AI interactions, we found that benefits in story originality were specifically predicted by engagement in idea generation with ChatGPT (prompting the AI to provide new story concepts, plot elements, or character ideas), but not idea elaboration (asking the AI to refine, reframe, or evaluate previously generated ideas). Additionally, we found that incorporating AI-generated content into the stories directly correlated with increased originality. Our findings suggest that AI-driven benefits depend on active engagement specifically focused on idea generation, although AI collaboration can limit idea diversity.

• Introduces two automated transformer models for assessing cognitive question complexity: high-performing LLaMA 2 and efficient, robust RoBERTa. • These fine-tuned transformer language models greatly outperform surface-based and semantic-distance baselines. • Both models predict open-ended question complexity and align strongly with human ratings. • Introduces PrediQT , a free web app for automated question-complexity scoring. • Bridges creativity, cognition, and AI for educational assessment innovation. Question-asking, an essential yet understudied activity, holds significant implications for learning, creativity, and cognitive development. Research shows that asking complex, open-ended questions is better for learning than closed ones. Previous research has explored open-ended question complexity through Bloom's taxonomy, but the measurement of complexity remains challenging. Recent advancements in natural language processing have enabled automated scoring of psychological tasks aligned to human-ratings. However, automatic assessment of open-ended questions remains understudied. We address this gap by fine-tuning transformer language models to predict human ratings of open-ended question complexity and comparing them to existing baseline measures (i.e., word count and semantic distance). Using previously collected human-rated responses and Bloom ratings from a creative question-asking task, we trained an encoder model (RoBERTa) and a Large Language Model (Llama-2-7B). Our results reveal that RoBERTa correlated strongly with human ratings of complexity (r = .73), exceeding baseline measures and offering an efficient, lightweight solution suitable for broad adoption. Our fine-tuned LLaMA 2 model achieved stronger performance (r = .84), establishing a new benchmark for predictive accuracy. Thus, we demonstrate how language models can be utilized to automatically score the complexity of open-ended questions. Importantly, LLaMA 2 demonstrates higher accuracy, while RoBERTa provides a replicable, accessible, and cost-effective option for everyday educational and psychological applications. Our work paves the way for automatic assessment of open-ended questions, which are critical across a wide range of cognitive domains.