United States

Conventional visual complexity measurement faces challenges in efficiency, accuracy, and alignment with human perception. To address these, this paper presents a novel Monte Carlo-based method for visual complexity measurement, the random line segment width sampling algorithm (RLSWSA). RLSWSA employs local stochastic sampling for efficient estimation of symbol perimeter complexity. By discarding global scanning in favor of local sampling, RLSWSA significantly enhances computational efficiency while maintaining high accuracy and robustness. Experimental results show rapid convergence with just 24 samples, yielding high consistency with traditional methods (correlation coefficient &gt; 0.9). Furthermore, RLSWSA&#39;s Spearman correlation with human subjective ratings is 0.74, demonstrating its strong correlation with human perceptual complexity. This study offers an efficient and reliable solution for rapid symbol visual complexity calculation, with strong potential for applications like symbol recognition and design optimization.

CogSci 2025

A Visual Complexity Measurement Method Based on Monte Carlo Sampling

aesthetics

computational modeling

computer science

psychology

representation

### Welcome to CogSci Conference 2025!

The 47th Annual Meeting of the Cognitive Science Society was a hybrid meeting held in San Francisco. 

<div style="position:relative;padding-top:0;width:900px;height:500px;"><iframe style="position:absolute;border:none;width:100%;height:100%;left:0;top:0;" src="https://online.fliphtml5.com/ebtyf/amvr/"  seamless="seamless" scrolling="no" frameborder="0" allowtransparency="true" allowfullscreen="true" ></iframe></div>

#### About

The Cognitive Science Society brings together researchers from around the world who hold a common goal: understanding the nature of the human mind. The mission of the Society is to promote Cognitive Science as a discipline, and to foster scientific interchange among researchers in various areas of study, including Artificial Intelligence, Linguistics, Anthropology, Psychology, Neuroscience, Philosophy, and Education.

The Society is a non-profit professional organization and its activities include sponsoring an annual conference and publishing the journals Cognitive Science and TopiCS.

#### Our History 

* **Society Creation**<br>
The Society was incorporated as a 501(c)(3) non-profit professional organization in Massachusetts in 1979. The organizing committee included Roger Schank, Allan Collins, Donald Norman, and a number of other scholars from psychology, linguistics, computer science, and philosophy. 
<br><br>
* **Conference Creation**<br>
The first conference on cognitive science was held at La Jolla, California in August, 1979, and has occurred annually since then. The proceedings of each conference are published, and those from most years are available through Lawrence Erlbaum Associates, Inc. The annual proceedings of the Cognitive Science Conference represent a major source of information on new work and new ideas in the scientific study of thinking. In 1990, the Society, with help from an anonymous donor, established the David Marr Prize for the best student paper at each annual meeting.
<br><br>
* **Journal Creation**<br>
The Journal, Cognitive Science, began publication in 1976, and is now published by Wiley-Blackwell. The Executive Editor is currently Richard P. Cooper of Birkbeck, University of London, and there are 18 Associate Editors and a 30-member editorial board. It serves as the premier outlet for research reports that intersect two or more disciplines. Copyrights for articles published in the journal are held by the Society. The Governing Board of the Cognitive Science Society voted in late 2006 to found a new journal, Topics in Cognitive Science (topiCS). The Editor in Chief is Wayne Gray, Cognitive Science Department, Rensselaer Polytechnic Institute. The journal seeks to fill a niche not occupied by Cognitive Science Journal or other cognitive science journals. Membership in the Society includes a subscription to Cognitive Science and TopiCS. Copyrights for articles published in the journal are held by the Society.
<br><br>

#### Code of Conduct

By attending the CogSci 2025 Conference, you are required to adhere to the society’s **[Code of Conduct](https://drive.google.com/file/d/1ChPuihLy6jE_BWqfO7J2KKgX35JW2zsM/view?usp=sharing)**.
<br><br>


You need to log in with the email address you registered with. 

Login credentials were sent to you from Underline -  subject line "Welcome to the CogSci 2025 Conference". Please be sure to check your spam/promotional inbox  if you do not see an email confirmation right away.





Please log in to join this event.

To access the site, please register [**here**](https://cognitivesciencesociety.org/registration/).

If you are registered and feel like you are seeing this message by mistake, please make sure you are logged in with the same email that you registered with. 

Please register!

The 47th Annual Meeting of the Cognitive Science Society presents the latest research across cognitive science and highlights the theme of Cognition in Context.

Conventional visual complexity measurement faces challenges in efficiency, accuracy, and alignment with human perception. To address these, this paper presents a novel Monte Carlo-based method for visual complexity measurement, the random line segment width sampling algorithm (RLSWSA). RLSWSA employs local stochastic sampling for efficient estimation of symbol perimeter complexity. By discarding global scanning in favor of local sampling, RLSWSA significantly enhances computational efficiency while maintaining high accuracy and robustness. Experimental results show rapid convergence with just 24 samples, yielding high consistency with traditional methods (correlation coefficient > 0.9). Furthermore, RLSWSA's Spearman correlation with human subjective ratings is 0.74, demonstrating its strong correlation with human perceptual complexity. This study offers an efficient and reliable solution for rapid symbol visual complexity calculation, with strong potential for applications like symbol recognition and design optimization.

poster

How to improve the persuasiveness of online requests is crucial to achieve acceptance and foster positive social relationships. The effectiveness of persuasion strategies and the influence of the sequence in which these strategies are applied have been demonstrated in the literature. However, existing research has largely overlooked the importance of linking the sequential persuasion strategies to the specific purpose. In this study, we first employ a few-shot Iterative Collaboration Method (ICM) to identify the purpose of the online requests, referencing human needs as well as the persuasion strategies used. Then, the sequential patterns of persuasion strategies supporting respective purposes are mined. Finally, Large Language Models (LLMs), incorporating the identified effective sequential strategies are used to rewrite the original requests. The results indicate that the sequence of strategies used for different purposes, can significantly increase the level of persuasion. The code and dataset can be found at https://github.com/phillip2f/Seq-Strategies-and-Purpose.

Can Sequential Persuasion Strategies Referencing Specific Purposes Enhance the Persuasiveness of Online Requests? A Case Study

We conducted a survey of over 2000 people before and after the Great East Japan Earthquake and obtained valid responses from 840 participants who completed both surveys. Participants responded to two types of trolley problems and answered a series of sociodemographic questions. As in previous studies, more than 80% of participants responded that they would push the switch (utilitarian judgment), and only less than 45% of participants chose the utilitarian option in the modified version (i.e., pushing to drop the person to stop the trolley). We found that people were slightly less likely to flip the switch after the earthquake than before. In addition, people who believed the world was trustworthy and who had a greater sense of self-control were more likely to flip. The results are discussed in terms of the flexibility and adaptability of external environmental factors and their possible effects on moral judgments.

Influences of catastrophic events on ethical judgments:  A case from the Great East Japan Earthquake

Heterogeneity in open fMRI datasets, caused by variations in scanning protocols, confounders, and population diversity, hinders representation learning and classification performance. To address these limitations, we propose a novel multi-site adversarial learning network (MSalNET) for fMRI-based mental disorder detection. Firstly, a representation learning module is introduced with a node information assembly (NIA) mechanism to extract features from functional connectivity (FC). This mechanism aggregates edge information from both horizontal and vertical directions, effectively assembling node information. Secondly, to generalize the feature across sites, we proposed a site-level feature extraction module that can learn from individual FC. Lastly, an adversarial learning network, is proposed to balance the trade-off between individual classification and site regression tasks. The proposed method was evaluated on Autism Brain Imaging Data Exchange (ABIDE). The results indicate that the proposed method achieves an accuracy of 75.56% with reducing variability from a data-driven perspective. The most discriminative brain regions revealed by NIA are consistent with statistical findings, uncovering the black box of deep learning to a certain extent. MSalNET offers a novel perspective on the detection of multi-site fMRI mental disorders and it considers the interpretability of the model, which is a crucial aspect in deep learning.

Multi-site fMRI-based mental disorder detection using adversarial learning: an ABIDE study

Knowledge Graph Reasoning (KGR) is an effective way to ameliorate sparsity and incompleteness problems by inferring new knowledge based on existing knowledge. The probabilistic case-based reasoning (CBR) model can gather reasoning paths from similar entities and relations in KG, thus outperforming rule-based and embedding-based KGR methods. However, it is still limited by some problems, such as insufficient learning of similarity features and sparse intermediate activations. This paper proposes a \textbf{S}imilarity-\textbf{A}ware graph transformer-enhanced probabilistic CBR model for \textbf{KGR}, namely SA-KGR. The proposed model regards the reasoning task as the KG query answering and is composed of two phases. The first phase is similarity-aware graph transformer-based graph feature encoding, which equips the similarity matrix and Mixture-of-Expert network to obtain fine-grained similarity features that are more helpful for reasoning path generation. The second phase is similarity-enhanced probabilistic case-based reasoning, which can retrieve and infer query answers from the generated candidate paths to complete brain-like cognitive reasoning. Extensive experimental results on various benchmarks unambiguously demonstrate that the proposed SA-KGR model can obtain the state-of-the-art results of current CBR-based methods.

A Similarity-Aware Graph Transformer-enhanced Probabilistic Case-based Reasoning Model for Knowledge Graph Reasoning

Temporal-spatial metaphors can be differed across languages, and such cross-linguistic influences may affect people’s spatial conceptualization of time. Mandarin (including gestures) has different spatial metaphors for time than Chinese Sign Language (CSL). This paper investigated whether native Mandarin speakers’ mental space-time mappings change after learning CSL for 14 weeks. Sixty native Mandarin speakers who had no prior knowledge of sign language took a pretest and posttest of space-time mappings before and after taking a CSL course. The results showed that participants changed their temporal-spatial mappings after learning CSL. Specifically, they had more sagittal space-time mappings and fewer lateral ones than before. They also had more “future-in-front/ past-in-back” mappings consistent with CSL space-time mappings. Furthermore, these changes were more significant in high-proficiency learners than in low-proficiency learners. Our results not only demonstrate an effect of bodily experience on time conceptions, but also suggest that sign language can impact spatial temporal reasoning.

The effect of learning Chinese Sign Language on spatial conceptualisation of time in hearing Mandarin speakers

Gaze Interaction Behavior Recognition (GIBR) plays a significant role in understanding social behaviors and diagnosing mental health conditions. However, existing methods are limited by inadequate task modeling, resulting in suboptimal performance. To address this issue, we model the GIBR task as a spatiotemporal reasoning problem integrating three modalities: emotion, gaze, and position. Based on this, we propose GIBRNet, which enhances the representation of gaze interaction tendencies through an Emotion-Aware Refinement Matrix and dynamically aggregates multi-frame, multi-modal information using GP-GNN, enabling more precise interaction behavior reasoning. Comparative experiments on the VACATION dataset demonstrate that GIBRNet significantly outperforms existing approaches. Additionally, we constructed a GIBR dataset suite, consisting of three extended datasets, for generalization evaluation, demonstrating GIBRNet's superiority. All datasets and code are publicly available.

GIBRNet: A Multimodal Spatiotemporal Reasoning Network Integrating Emotion, Gaze, and Position for Gaze Interaction Behavior Recognition

Understanding how psychological constructs are formalized into measurable representations remains a central challenge in cognitive science. This study examines large language models (LLMs) as cognitive artifacts capable of decomposing and formalizing such constructs. We evaluate LLMs’ conceptual alignment with human cognition across three tasks: predicting psychological attributes from text, generating theoretical construct explanations, and creating psychometrically valid assessment items. Results show strong representational fidelity for personality traits (average r=0.582), moderate alignment for depression (r=0.515), and emerging capabilities for anxiety (r =0.259). Expert evaluations confirm that LLMs produce theoretically coherent construct explanations (mean rating=4.33/5), while generated questionnaires exhibit convergent validity with established measures. Our findings position LLMs as effective tools for exploring cognitive representation systems and suggest that theory-driven prompting enables reliable, automated questionnaire development. This work bridges AI formalization techniques with core questions in cognitive science, offering a scalable framework for studying and constructing psychological measurement tools.

Cognitive Representation in Large Language Models: Formalizing Psychological Constructs for Automated Questionnaire Generation

Large language models (LLMs) excel at personality prediction but are often impractical for deployment due to high computational demands. This work introduces cognitive distillation with Chain-of-Thought calibration, a novel framework for transferring structured reasoning from large LLMs to smaller, efficient models. Inspired by cognitive architectures like ACT-R, our method aligns intermediate inference steps using exemplars from single or multiple LLMs. A 1.5B-parameter model distilled through this process surpasses Qwen1.5-110B in predictive accuracy, achieving a 28% improvement in Pearson correlation while using just 1.36% of its parameters. Ablation studies reveal that moderate sampling diversity and multi-model ensembles enhance cognitive skill transfer and construct validity. These findings demonstrate that high-level reasoning can be efficiently and faithfully transferred, enabling psychometrically robust personality assessment in resource-constrained settings. This approach bridges AI and cognitive science, offering a scalable path toward plausible, cognitively grounded language models.

Cognitive Distillation with Parameter-Efficient LLMs: Chain-of-Thought Calibration for Personality Prediction

As Large Language Models (LLMs) become more prevalent, concerns about their safety, ethics, and potential biases have risen. Systematically evaluating LLMs' risk decision-making tendencies, particularly in the ethical domain, has become crucial. This study innovatively applies the Domain-Specific Risk-Taking (DOSPERT) scale from cognitive science to LLMs and proposes a novel Ethical Decision-Making Risk Attitude Scale (EDRAS) to assess LLMs' ethical risk attitudes in depth. We further propose a novel approach integrating risk scales and role-playing to quantitatively evaluate systematic biases in LLMs. Through systematic evaluation of multiple mainstream LLMs, we assessed the "risk personalities" of LLMs across multiple domains, with a particular focus on the ethical domain, and revealed and quantified LLMs' systematic biases towards different groups. This helps understand LLMs' risk decision-making and ensure their safe and reliable application. Our approach provides a tool for identifying biases, contributing to fairer and more trustworthy AI systems.

Quantifying Risk Propensities of Large Language Models: Ethical Focus and Bias Detection through Role-Play

The emergence of large language model (LLM) has promoted the research progress in many fields, but it still faces challenges in imitating human logical reasoning, especially in the step-by-step reasoning of complex tasks and zero-shot logical cognition. To address these challenges, we propose a multi-step symbolic reasoning strategy that decomposes complex tasks into subtasks and optimizes the decomposition using a subtask verification module. Moreover, we also introduce a new zero-shot symbolic module which can help improve the model’s reasoning ability on unseen samples with symbolic representation and logical schemes. We evaluated our method on four reasoning datasets: the industrial private dataset Ship Assembly Technology and the public datasets ProntoQA, ProofWriter, and OpenBookQA. Our framework demonstrates substantial improvements in reasoning interpretability and generalization capacity compared to existing prompting paradigms. The proposed method establishes a new pathway for enhancing LLMs' cognitive architectures through symbolic system integration, showing strong potential for efficient knowledge transfer to downstream applications while preserving human-understandable reasoning traces.

Downloads

Next from CogSci 2025

Can Sequential Persuasion Strategies Referencing Specific Purposes Enhance the Persuasiveness of Online Requests? A Case Study

Similar lecture

Complexity in Complexity: Understanding Visual Complexity Through Structure, Color, and Surprise