Hasil untuk "Organizational behaviour, change and effectiveness. Corporate culture"

Rui Liu

Can an expensive AI model effectively direct a cheap one to solve software engineering tasks? We study this question by introducing ManagerWorker, a two-agent pipeline where an expensive "manager" model (text-only, no code execution) analyzes issues, dispatches exploration tasks, and reviews implementations, while a cheap "worker" model (with full repo access) executes code changes. We evaluate on 200 instances from SWE-bench Lite across five configurations that vary the manager-worker relationship, pipeline complexity, and model pairing. Our findings reveal both the promise and the limits of multi-agent direction: (1) a strong manager directing a weak worker (62%) matches a strong single agent (60%) at a fraction of the strong-model token usage, showing that expensive reasoning can substitute for expensive execution; (2) a weak manager directing a weak worker (42%) performs worse than the weak agent alone (44%), demonstrating that the directing relationship requires a genuine capability gap--structure without substance is pure overhead; (3) the manager's value lies in directing, not merely reviewing--a minimal review-only loop adds just 2pp over the baseline, while structured exploration and planning add 11pp, showing that active direction is what makes the capability gap productive; and (4) these behaviors trace to a single root cause: current models are trained as monolithic agents, and splitting them into director/worker roles fights their training distribution. The pipeline succeeds by designing around this mismatch--keeping each model close to its trained mode (text generation for the manager, tool use for the worker) and externalizing organizational structure to code. This diagnosis points to concrete training gaps: delegation, scoped execution, and mode switching are skills absent from current training data.

Shreya Havaldar, Sunny Rai, Young-Min Cho et al.

Existing benchmarks that measure cultural adaptation in LLMs are misaligned with the actual challenges these models face when interacting with users from diverse cultural backgrounds. In this work, we introduce the first framework and benchmark designed to evaluate LLMs in realistic, multicultural conversational settings. Grounded in sociocultural theory, our framework formalizes how linguistic style - a key element of cultural communication - is shaped by situational, relational, and cultural context. We construct a benchmark dataset based on this framework, annotated by culturally diverse raters, and propose a new set of desiderata for cross-cultural evaluation in NLP: conversational framing, stylistic sensitivity, and subjective correctness. We evaluate today's top LLMs on our benchmark and show that these models struggle with cultural adaptation in a conversational setting.

Florian Schneider, Carolin Holtermann, Chris Biemann et al.

Large Vision-Language Models (LVLMs) have recently gained attention due to their distinctive performance and broad applicability. While it has been previously shown that their efficacy in usage scenarios involving non-Western contexts falls short, existing studies are limited in scope, covering just a narrow range of cultures, focusing exclusively on a small number of cultural aspects, or evaluating a limited selection of models on a single task only. Towards globally inclusive LVLM research, we introduce GIMMICK, an extensive multimodal benchmark designed to assess a broad spectrum of cultural knowledge across 144 countries representing six global macro-regions. GIMMICK comprises six tasks built upon three new datasets that span 728 unique cultural events or facets on which we evaluated 20 LVLMs and 11 LLMs, including five proprietary and 26 open-weight models of all sizes. We systematically examine (1) regional cultural biases, (2) the influence of model size, (3) input modalities, and (4) external cues. Our analyses reveal strong biases toward Western cultures across models and tasks and highlight strong correlations between model size and performance, as well as the effectiveness of multimodal input and external geographic cues. We further find that models have more knowledge of tangible than intangible aspects (e.g., food vs. rituals) and that they excel in recognizing broad cultural origins but struggle with a more nuanced understanding.

Jean de Dieu Nyandwi, Yueqi Song, Simran Khanuja et al.

Multimodal Large Language Models excel in high-resource settings, but often misinterpret long-tail cultural entities and underperform in low-resource languages. To address this gap, we propose a data-centric approach that directly grounds MLLMs in cultural knowledge. Leveraging a large scale knowledge graph from Wikidata, we collect images that represent culturally significant entities, and generate synthetic multilingual visual question answering data. The resulting dataset, CulturalGround, comprises 22 million high-quality, culturally-rich VQA pairs spanning 42 countries and 39 languages. We train an open-source MLLM CulturalPangea on CulturalGround, interleaving standard multilingual instruction-tuning data to preserve general abilities. CulturalPangea achieves state-of-the-art performance among open models on various culture-focused multilingual multimodal benchmarks, outperforming prior models by an average of 5.0 without degrading results on mainstream vision-language tasks. Our findings show that our targeted, culturally grounded approach could substantially narrow the cultural gap in MLLMs and offer a practical path towards globally inclusive multimodal systems.

Pablo Jaramillo, Ivan Sipiran

This article explores the use of recent generative AI algorithms for repairing cultural heritage objects, leveraging a conditional diffusion model designed to reconstruct 3D point clouds effectively. Our study evaluates the model's performance across general and cultural heritage-specific settings. Results indicate that, with considerations for object variability, the diffusion model can accurately reproduce cultural heritage geometries. Despite encountering challenges like data diversity and outlier sensitivity, the model demonstrates significant potential in artifact restoration research. This work lays groundwork for advancing restoration methodologies for ancient artifacts using AI technologies.

Millicent Ochieng, Varun Gumma, Sunayana Sitaram et al.

The deployment of Large Language Models (LLMs) in real-world applications presents both opportunities and challenges, particularly in multilingual and code-mixed communication settings. This research evaluates the performance of seven leading LLMs in sentiment analysis on a dataset derived from multilingual and code-mixed WhatsApp chats, including Swahili, English and Sheng. Our evaluation includes both quantitative analysis using metrics like F1 score and qualitative assessment of LLMs' explanations for their predictions. We find that, while Mistral-7b and Mixtral-8x7b achieved high F1 scores, they and other LLMs such as GPT-3.5-Turbo, Llama-2-70b, and Gemma-7b struggled with understanding linguistic and contextual nuances, as well as lack of transparency in their decision-making process as observed from their explanations. In contrast, GPT-4 and GPT-4-Turbo excelled in grasping diverse linguistic inputs and managing various contextual information, demonstrating high consistency with human alignment and transparency in their decision-making process. The LLMs however, encountered difficulties in incorporating cultural nuance especially in non-English settings with GPT-4s doing so inconsistently. The findings emphasize the necessity of continuous improvement of LLMs to effectively tackle the challenges of culturally nuanced, low-resource real-world settings and the need for developing evaluation benchmarks for capturing these issues.

Kira Sam

Bacteraemia, a bloodstream infection with high morbidity and mortality rates, poses significant diagnostic challenges. Accurate diagnosis through blood cultures is resource-intensive. Developing a machine learning model to predict blood culture outcomes in emergency departments offers potential for improved diagnosis, reduced healthcare costs, and mitigated antibiotic use.This thesis aims to identify optimal machine learning techniques for predicting bacteraemia and develop a predictive model using data from St. Antonius Hospital's emergency department. Based on current literature, CatBoost and Random Forest were selected as the most promising techniques. Model optimization using Optuna prioritized sensitivity.The final Random Forest model achieved an ROC AUC of 0.78 and demonstrated 0.92 sensitivity on the test set. Notably, it accurately identified 36.02% of patients at low risk of bacteraemia, with only 0.85% false negatives.Implementation of this model in St. Antonius Hospital's emergency department could reduce blood cultures, healthcare costs, and antibiotic treatments. Future studies should focus on external validation, exploring advanced techniques, and addressing potential confounders to ensure model generalizability.

Yushi Yang, Francesco Turci, Erika Kague et al.

Collective behaviour in living systems is observed across many scales, from bacteria to insects, to fish shoals. Zebrafish have emerged as a model system amenable to laboratory study. Here we report a three-dimensional study of the collective dynamics of fifty zebrafish. We observed the emergence of collective behaviour changing between \yy{ordered} to randomised, upon \yy{adaptation} to new environmental conditions. We quantify the spatial and temporal correlation functions of the fish and identify two length scales, the persistence length and the nearest neighbour distance, that capture the essence of the behavioural changes. The ratio of the two length scales correlates robustly with the polarisation of collective motion that we explain with a reductionist model of self--propelled particles with alignment interactions.

Alexander J McGhee, Eric O McGhee, Jack E Famiglietti et al.

Advanced cell culture techniques such as 3D bio-printing and hydrogel-based cell embedding techniques harbor many new and exciting opportunities to study cells in environments that closely recapitulate in-vivo conditions. Researchers often study these environments using fluorescence microscopy to visualize the protein association with objects such as cells within the 3D environment, yet quantification of concentration profiles in the microenvironment has remained elusive. Here, we present a method to continuously measure the time-dependent concentration gradient of various biomarkers within a 3D cell culture assay using bead-based immunoassays to sequester and concentrate the fluorescence intensity of these tagged proteins. This assay allows for near real-time in situ biomarker detection and enables spatiotemporal quantification of biomarker concentration. Snapshots of concentration profiles can be taken, or time series analysis can be performed enabling time-varying biomarker production estimation. Example assays utilize an osteosacroma tumoroid as a case study for a quantitative single-plexed gel encapsulated assay, and a qualitative multi-plexed 3D bioprinted assay. In both cases, a time-varying cytokine concentration gradient is measured. An estimation for the production rate of the IL-8 cytokine per second per osteosarcoma cell results from fitting an analytical function for continuous point source diffusion to the measured concentration gradient and reveals that each cell produces approximately two IL-8 cytokines per second. Proper calibration and use of this assay is exhaustively explored for the case of diffusion-limited Langmuir kinetics of a spherical adsorber.

Valerii Semenets, Svitlana Gryshko, Mariia Golovianko et al.

Information retrieval (IR) is known facilitator of changes ongoing in human society and vice versa. This is due to the fact that IR is a key component of the digital ecosystems, where both information providers and information consumers collaboratively address their problems with the use of technologies. Organization and design of such ecosystems drives particular social impact for all the players involved. In this paper, we study the impact made by a particular IR ecosystem (semantic portal) used for management of academic information resources and processes within the Ukrainian higher education. We show how this portal is changing a collective mindset of the academic community of its users. We argue that such impact becomes possible due to specific organization of the ecosystem, where all the information resources, IR services and related analytics (search, assessment, ranking, etc.) and IR users inhabit the same semantic space under umbrella of the corresponding ontologies. Personal values and preferences of the users configure on-the-fly the corresponding IR analytics and enable personalized value-driven IR services, making everyone feel involved into the organizational decision-making processes. Four years of active use of this portal in university environment has been reported and related impact is evaluated in this study.

Tao Li, Quan Qiu, Chunjiang Zhao

The paper is concerned with the dynamic tracking problem of SNAP orchards harvesting robots in the presence of multiple uncalibrated model parameters in the application of dwarf culture orchards harvest. A new hybrid visual servoing adaptive tracking controller and three adaptive laws are proposed to guarantee harvesting robots to finish the dynamic harvesting task and the adaption to unknown parameters including camera intrinsic and extrinsic model and robot dynamics. By the Lyapunov theory, asymptotic convergence of the closed-loop system with the proposed control scheme is rigorously proven. Experimental and simulation results have been conducted to verify the performance of the proposed control scheme. The results demonstrate its effectiveness and superiority.

Dustin S. Stoltz, Marshall A. Taylor

Using the frequency of keywords is a classic approach in the formal analysis of text, but has the drawback of glossing over the relationality of word meanings. Word embedding models overcome this problem by constructing a standardized and continuous "meaning space" where words are assigned a location based on relations of similarity to other words based on how they are used in natural language samples. We show how word embeddings are commensurate with prevailing theories of meaning in sociology and can be put to the task of interpretation via two kinds of navigation. First, one can hold terms constant and measure how the embedding space moves around them--much like astronomers measured the changing of celestial bodies with the seasons. Second, one can also hold the embedding space constant and see how documents or authors move relative to it--just as ships use the stars on a given night to determine their location. Using the empirical case of immigration discourse in the United States, we demonstrate the merits of these two broad strategies for advancing important topics in cultural theory, including social marking, media fields, echo chambers, and cultural diffusion and change more broadly.

Subhadeep, Mukhopadhyay, Kaijun Wang

In a landmark paper published in 2001, Leo Breiman described the tense standoff between two cultures of data modeling: parametric statistical and algorithmic machine learning. The cultural division between these two statistical learning frameworks has been growing at a steady pace in recent years. What is the way forward? It has become blatantly obvious that this widening gap between "the two cultures" cannot be averted unless we find a way to blend them into a coherent whole. This article presents a solution by establishing a link between the two cultures. Through examples, we describe the challenges and potential gains of this new integrated statistical thinking.

Thomas M McGrath, Eleanor Spreckley, Aina Fernandez Rodriguez et al.

Better understanding of feeding behaviour will be vital in reducing obesity and metabolic syndrome, but we lack a standard model that captures the complexity of feeding behaviour. We construct an accurate stochastic model of rodent feeding at the bout level in order to perform quantitative behavioural analysis. Analysing the different effects on feeding behaviour of PYY 3-36, lithium chloride, GLP-1 and leptin shows the precise behavioural changes caused by each anorectic agent, and demonstrates that these changes do not mimic satiety. In the ad libitum fed state during the light period, meal initiation is governed by complete stomach emptying, whereas in all other conditions there is a graduated response. We show how robust homeostatic control of feeding thwarts attempts to reduce food intake, and how this might be overcome. In silico experiments suggest that introducing a minimum intermeal interval or modulating gastric emptying can be as effective as anorectic drug administration.

Niklas Stoehr, Fabian Braesemann, Michael Frommelt et al.

The digital transformation is driving revolutionary innovations and new market entrants threaten established sectors of the economy such as the automotive industry. Following the need for monitoring shifting industries, we present a network-centred analysis of car manufacturer web pages. Solely exploiting publicly-available information, we construct large networks from web pages and hyperlinks. The network properties disclose the internal corporate positioning of the three largest automotive manufacturers, Toyota, Volkswagen and Hyundai with respect to innovative trends and their international outlook. We tag web pages concerned with topics like e-mobility and environment or autonomous driving, and investigate their relevance in the network. Sentiment analysis on individual web pages uncovers a relationship between page linking and use of positive language, particularly with respect to innovative trends. Web pages of the same country domain form clusters of different size in the network that reveal strong correlations with sales market orientation. Our approach maintains the web content's hierarchical structure imposed by the web page networks. It, thus, presents a method to reveal hierarchical structures of unstructured text content obtained from web scraping. It is highly transparent, reproducible and data driven, and could be used to gain complementary insights into innovative strategies of firms and competitive landscapes, which would not be detectable by the analysis of web content alone.

Behnam Malmir, Shing I Chang

This paper describes a method to convert Microsoft Kinect coordinates into gait parameters in order to detect a person's gait change. The proposed method can help quantify the progress of physical therapy. Microsoft Kinect, a popular platform for video games, was used to generate 25 joints to form a human skeleton, and then the proposed method converted the coordinates of selected Kinect joints into gait parameters such as spine tilt, hip tilt, and shoulder tilt, which were tracked over time. Sample entropy measure was then applied to quantify the variability of each gait parameter. Male and female subjects walked a three-meter path multiple times in initial experiments, and their walking patterns were recorded via the proposed Kinect device through the frontal plane. Time series of the gait parameters were generated for subjects with and without knee braces. Sample entropy was used to transform these time series into numerical values for comparison of these two conditions.

D. K. Watson

Understanding the emergence of collective organizational phenomena is a major goal in many fields of physics from condensed matter to cosmology. Using a recently introduced manybody perturbation formalism for fermions, we propose a mechanism for the emergence of collective behavior, specifically superfluidity, driven by quantum statistics and the enforcement of the Pauli principle through the selection of normal modes. The method, which is called symmetry invariant perturbation theory (SPT), uses group theory and graphical techniques to solve the manybody Schrodinger equation through first order exactly. The solution at first order defines collective coordinates in terms of five N-body normal modes, identified as breathing, center of mass, single particle angular excitation, single particle radial excitation and phonon. A correspondence is established "on paper" that enforces the Pauli principle through the assignment of specific normal mode quantum numbers. Applied in the unitary regime, this normal mode assignment yields occupation only in an extremely low frequency N-body phonon mode at ultralow temperatures. A single particle radial excitation mode at a much higher frequency creates a gap that stabilizes the superfluidity at low temperatures. Coupled with the corresponding values for the frequencies at unitarity obtained by this manybody calculation, we obtain good agreement with experimental thermodynamic results including the lambda transition in the specific heat. Our results suggest that the emergence of collective behavior in macroscopic systems is driven by the Pauli principle and its selection of the correct collective coordinates in the form of N-body normal modes.

Mark E. Madsen, Carl P. Lipo

Cultural transmission models are coming to the fore in explaining increases in the Paleolithic toolkit richness and diversity. During the later Paleolithic, technologies increase not only in terms of diversity but also in their complexity and interdependence. As Mesoudi and O'Brien (2008) have shown, selection broadly favors social learning of information that is hierarchical and structured, and multiple studies have demonstrated that teaching within a social learning environment can increase fitness. We believe that teaching also provides the scaffolding for transmission of more complex cultural traits. Here, we introduce an extension of the Axelrod (1997} model of cultural differentiation in which traits have prerequisite relationships, and where social learning is dependent upon the ordering of those prerequisites. We examine the resulting structure of cultural repertoires as learning environments range from largely unstructured imitation, to structured teaching of necessary prerequisites, and we find that in combination with individual learning and innovation, high probabilities of teaching prerequisites leads to richer cultural repertoires. Our results point to ways in which we can build more comprehensive explanations of the archaeological record of the Paleolithic as well as other cases of technological change.

Su Dong, Paul Heroux

A typical cell culture CO2 incubator was probed in detail to document the pattern of 60-Hz magnetic fields (MFs) inside the unit, as well as the ability of the incubator to attenuate environmental MFs. Subsequently, a survey of 46 cell culture incubators was performed. The survey measured MFs outside and inside the incubators, the frequency spectrum between 5 and 2000 Hz, and variations over time of the 60-Hz MF. Our measurements show an uneven spatial distribution, reflecting electronic and electrical components hidden within the walls. Attenuation of environmental MFs varied between 18 % and 33 %, signalling easy penetration into the units. MF levels, frequency spectra and variations over time were very different from one unit to the next. All 46 incubators surveyed had an average field greater than 0.2 microT; among them, 39 (85 %) had an average field greater than 1 microT. There is substantial work to be done in improving control over the MF environment of in vitro experiments in bio-medicine, particularly if they involve cancer cells.

A. Sarkar, P. Barat

The plastic deformation of dilute alloys is often accompanied by plastic instabilities due to dynamic strain aging and dislocation interaction. The repeated breakaway of dislocations from and their recapture by solute atoms leads to stress serrations and localized strain in the strain controlled tensile tests, known as the Portevin-Le Chatelier (PLC) effect. In this present work, we analyse the stress time series data of the observed PLC effect in the constant strain rate tensile tests on Al-2.5%Mg alloy for a wide range of strain rates at room temperature. The scaling behaviour of the PLC effect was studied using two complementary scaling analysis methods: the finite variance scaling method and the diffusion entropy analysis. From these analyses we could establish that in the entire span of strain rates, PLC effect showed Levy walk property. Moreover, the multiscale entropy analysis is carried out on the stress time series data observed during the PLC effect to quantify the complexity of the distinct spatiotemporal dynamical regimes. It is shown that for the static type C band, the entropy is very low for all the scales compared to the hopping type B and the propagating type A bands. The results are interpreted considering the time and length scales relevant to the effect.