Shared vocabularies are not designed by any central authority; they emerge from the bottom up as communities of speakers repeatedly try to understand one another. We study this process with an agent-based Naming Game in which a population that begins with no common vocabulary self-organises a shared language through local pairwise interaction, and we ask how the social network governs the outcome under communication noise and under selection on communicative strategy. The study combines baseline experiments with an extended suite built specifically to meet the standards of replication and statistical inference: every reported quantity is averaged over many independent seeded runs with confidence intervals, and the central claim is tested with a formal interaction statistic. Convergence time scales with population size as a power law with exponent close to four-thirds, matching analytical results for the Naming Game on networks. Restricted networks sustain many coexisting regional dialects, which we quantify with a cluster-based metric and show to be metastable, decaying to a single language as the interaction budget grows. Our central methodological contribution concerns noise robustness. A naive comparison suggests that denser networks simply tolerate more noise, but by comparing topologies at matched mean degree and by sweeping degree directly we separate two effects: connectivity raises robustness with sharply diminishing returns, while at identical degree locally clustered topologies remain significantly less robust than well-mixed ones. The effect of topology therefore survives the removal of the degree confound. Finally, an evolutionary extension shows that selection for communicative success drives inventiveness towards a low level that a frequency-dependent reward for useful innovation raises only slightly, so conservatism is robust rather than an artefact of the fitness function.

An agent-based model of one community whose members meet every week. Members decide whether to attend the community’s gatherings, meet partners there, reveal or withhold their views on two topics, and judge each partner as “like me” or as one of “those people”. Views move toward accepted voices and are held by each member’s starting view. Each conversation has a feeling, which updates the member’s warmth toward her own side and toward the other side and strengthens or weakens the tie between the two partners; ties decide belonging, and belonging decides attendance. Listening enters as one trait, conversational receptiveness. The release includes the BehaviorSpace setup file and the output of every run reported in the accompanying paper.

This study separates operational interoperability, fare integration and revenue-sharing governance in a stylised multi-operator bus network. An agent-based model implemented in NetLogo 7.0.3 represents passengers, buses and stations across 14 stations, nine routes and three operators. A 2 × 2 institutional design is complemented by financial clearing comparisons and 43,200 robustness runs: behavioural sensitivity (E1), a 243-configuration operational factorial (E2), and targeted joint behavioural–operational stress testing (E3). The audited experiment records show favourable interoperability effects on waiting, total travel time, transfers and realised generalised cost in all 1,944 E2 and 768 E3 configuration-level primary comparisons. Fare integration substantially reduces passenger expenditure but produces an aggregate deficit under the assumed accounting coefficients. Revenue-sharing rules redistribute operator-level outcomes without changing aggregate profit or passenger trajectories by construction. The model is intended to identify mechanisms and test conditional robustness, not predict city-specific policy effects. The paper contributes an explicit separation of access, pricing and settlement mechanisms coupled with reproducible factorial and stress-test comparisons.

This Python agent-based model simulates the diffusion of instructional practices among primary-school science teachers connected through within-school and cross-school networks. It represents teachers as agents nested in schools and compares five adoption mechanisms: simple contagion, proportion thresholds, count thresholds, adaptive norm conformity, and an individual-only baseline. The model evaluates fixed-budget training allocations, including concentrated and distributed policies, and estimates when school-level dose-response evidence can support targeted allocation. Experiments examine peer interference, cross-school contact intensity, global critical transitions, allocation stability, and the distinction between participant-level and system-level outcomes. Teacher attributes are informed by a de-identified Beijing questionnaire, whereas team sizes and contact networks are explicit model assumptions. The package includes executable Python scripts, ODD documentation, parameter tables, experiment metadata, simulation outputs, and the authorized questionnaire workbook required to reproduce the questionnaire-informed analyses. Results are intended as model-based design evidence for future longitudinal research rather than forecasts of adoption rates.

This model extends Schelling’s segregation model with mobility sharing. Agents of two groups occupy cells of a 32 × 32 grid and repeatedly choose activity destinations subject to a preference threshold, the minimum share of ingroup neighbours they accept. Trips are matched into shared rides between members of different groups, using one of four matching algorithms on a shareability network (maximal cross-group matching, maximum weight matching on contact time, maximum cardinality matching, or random greedy matching) with a maximum detour constraint. After each trip, an agent’s preference threshold changes with the time spent in contact with outgroup members.

Starting from complete segregation, the model examines under which conditions contact during shared trips leads to integration. Besides the main scenario, in which contact always lowers the threshold, it includes two extensions. In the first, a share of adversarial agents becomes less tolerant with contact, and this share grows or shrinks with the current level of segregation, governed by a branching dissimilarity index and a conversion rate. In the second, each contact raises or lowers the threshold with a probability that depends on the agent’s current attitude relative to a compatibility threshold.

Segregation is measured with the index of dissimilarity and the Gini coefficient. The model file includes BehaviorSpace experiments for each analysis in the associated article. The matching step uses the NetworkX library through the Python extension of NetLogo (requires Python 3 with numpy and networkx 3.x).

This repository contains a Python simulation script for studying network-based linguistic diversity. The model extends the classic Greenberg Diversity Index by introducing diversity measures defined over social ties rather than only over population shares. It also includes two extensions: a competence-conditioned version and a multilayer version.

Replication package for a network diffusion model of how Protestant belief spreads through a
signed social network, used to compare three versions of balance theory.

We built a model of knowledges diffusion in networks. The particularity is that knowledges to be passed on are cumulative : rely on previous acquisition of specific knowledges of a lower level to be passed from an agent to another. We test for different types of networks, different selection rules for the knowledge to be passed and the possibility to introduce a turnover among agents.
We concentrate on the learning speed and the convergence level of the learning process.

GFN & Technology Models Library (2025–2035) is a collection of four interrelated computational models focused on global financial networks, cascading default risks, technological development, and technological lag, with particular attention to Russia, the USA, and China. The library combines agent-based modeling, system dynamics, Monte-Carlo simulation, and network analysis. It includes: (1) a Numba-accelerated Monte-Carlo technology index model; (2) a hybrid GFN + cascading defaults (Gai–Kapadia) + Russian system dynamics model; (3) a twin-simulation GFN+SBS framework with selective bailout and Russia’s peripheral position; and (4) an endogenous GFN model with technology centrality, dependence, and tech lag dynamics under sanctions scenarios. GFN & Technology Models Library (2025–2035) is a collection of four interrelated computational models focused on global financial networks, cascading default risks, technological development, and technological lag, with particular attention to Russia, the USA, and China.

The library combines agent-based modeling, system dynamics, Monte-Carlo simulation, and network analysis. It includes:

1. Technology Index 2035 (Monte-Carlo + Numba) — hybrid SD/ABM model projecting a composite technology index under different scenarios (Russia baseline, China, USA) with aggressive Numba optimization.
2. RUS-GFN-TSI-2035 CASCADE — hybrid model integrating a Gai–Kapadia-style cascading defaults network, an agent-based Global Financial Network layer, and a system-dynamics module for the Russian economy (TSI, inflation, technology, trust, etc.).
   …

Organizations operate under conditions of imperfect performance in which human error is inevitable, yet errors are rarely examined as the triggers for the managerial interactions that shape organizational culture over time. The present research introduces an agent-based simulation of a work team completing a fixed sequence of tasks under varying degrees of managerial oversight and response policy. The model isolates the mechanical loss of throughput caused by errors from the psychological and cultural consequences of managerial reactions, which are categorized into ignoring, correcting or punishing. Furthermore, the model incorporates an autonomous self-notice mechanism, allowing workers to correct themselves in the absence of managerial intervention. By tracking the accumulation of worker resentment and the transient enhancements in learning, the simulation acts as a dynamic laboratory for observing delayed consequences, nonlinear tipping points and systemic organizational collapse. The results reveal a central paradox of organizational control. Highly monitored punitive environments generate high short-term throughput, yet they simultaneously accumulate interactional injustice and resentment that engineer a rapid cascading turnover and the highest probability of systemic collapse. Conversely, corrective policies combined with active monitoring achieve equivalent throughput while sustaining workforce viability. A comprehensive sensitivity analysis establishes that error accumulation is primarily determined by structural factors, namely agent-level mistake propensity and task difficulty, while resignation dynamics, resentment accumulation and collapse timing remain predominantly governed by managerial policy, a hierarchy independently corroborated by a surrogate model and shown to be stable across independent seeds and across stakeholder weighting scenarios. The study bridges the operational mechanics of task completion and the social dynamics of workplace mistreatment, illustrating how short-term punitive success often masks long-term structural fragility.