Abstract:Abstract Collective dynamics emerge from countless individual decisions. Yet, we poorly understand the processes governing dynamically-interacting individuals in human collectives under realistic conditions. We present a naturalistic immersive-reality experiment where groups of participants searched for rewards in different environments, studying how individuals weigh personal and social information and how this shapes individual and collective outcomes. Capturing high-resolution visual-spatial data, behavioral analyses revealed individual-level gains—but group-level losses—of high social information use and spatial proximity in environments with concentrated (vs. distributed) resources. Incentivizing participants at the group (vs. individual) level facilitated adaptation to concentrated environments, buffering apparently excessive scrounging. To infer discrete choices from unconstrained interactions and uncover the underlying decision mechanisms, we developed an unsupervised Social Hidden Markov Decision model. Computational results showed that participants were more sensitive to social information in concentrated environments frequently switching to a social relocation state where they approach successful group members. Group-level incentives reduced participants’ overall responsiveness to social information and promoted higher selectivity over time. Finally, mapping group-level spatio-temporal dynamics through time-lagged regressions revealed a collective exploration-exploitation trade-off across different timescales. Our study unravels the processes linking individual-level strategies to emerging collective dynamics, and provides tools to investigate decision-making in freely-interacting collectives.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: **How to reduce the over - exploitation problem of human groups in the unrestricted use of social information through collective incentives?** Specifically, the authors focus on how individuals in human groups weigh the use of personal and social information under natural conditions, and how this trade - off affects the performance of individuals and the collective. The focus of the study is to understand how individuals dynamically adjust their decision - making strategies under different environments and incentive structures to avoid the decline in collective performance caused by excessive reliance on social information. ### Main problems: 1. **Excessive reliance on social information**: In an environment where resources are centrally distributed, individuals tend to rely too much on resources discovered by others, resulting in insufficient exploration and over - exploitation of the environment. 2. **Conflict between individual and collective interests**: Individual incentives (i.e., paying rewards based on personal success) may lead individuals to rely too much on social information, thus harming the overall performance of the collective; while collective incentives (i.e., paying rewards based on collective success) may promote more efficient resource allocation and utilization. ### Research methods: - **Immersive reality experiment**: Participants search for resources ("coins") in a 3D virtual environment and observe the actions of other participants in real - time. - **High - resolution data collection**: Record the visual information and movement trajectories of each participant. - **Social Hidden Markov Decision model**: Used to infer hidden decision - making states (such as "individual exploration" or "social re - positioning") and describe the transition probabilities between these states. ### Key findings: - **Environmental impact**: In an environment where resources are concentrated, individuals are more likely to rely on social information, but the opposite is true in a distributed environment. - **Impact of incentive structures**: Collective incentives reduce participants' reliance on social information, promote higher selectivity, and partially alleviate the negative impacts brought by excessive reliance on social information. - **Dynamic changes on the time scale**: Over time, participants in resource - concentrated environments gradually become more selective in using social information, while remaining relatively stable in distributed environments. ### Conclusion: This study reveals how individual - level decision - making affects collective behavior and provides tools to study the decision - making process in freely interacting collectives. By introducing collective incentives, the decline in collective performance caused by excessive reliance on social information can be effectively reduced, thus achieving better resource allocation and utilization. ### Formula examples: The formulas involved in the paper include: \[ P_{I \rightarrow S}(t)=\text{logit}^{-1}\left(\alpha_{ij}[t]+\beta_V V_t+\beta_D D_t+\beta_N N_t+\beta_T T_t\right) \] where: - \( P_{I \rightarrow S}(t) \) represents the probability of switching from the individual exploration state to the social re - positioning state. - \( V_t \) is the number of visible mining players. - \( D_t \) is the distance of visible resource patches. - \( N_t \) is the number of visible mining players. - \( T_t \) is the time since the last success. These formulas help quantify the influence of different factors on individual decision - making, thereby revealing the mechanisms behind collective behavior.

Collective incentives reduce over-exploitation of social information in unconstrained human groups

Collective decision-making under changing social environments among agents adapted to sparse connectivity

Flexible social inference facilitates targeted social learning when rewards are not observable

Incentivizing free riders improves collective intelligence in social dilemmas

Agents seeking long-term access to the wisdom of the crowd reduce immediate decision-making accuracy

Strategic formation of collaborative networks

Information socialtaxis and efficient collective behavior emerging in groups of information-seeking agents

Information Synergy Maximizes the Growth Rate of Heterogeneous Groups

Does Spending More Always Ensure Higher Cooperation? An Analysis of Institutional Incentives on Heterogeneous Networks

Confidence over competence: Real-time integration of social information in human continuous perceptual decision-making

A Cognitive Computational Approach to Social and Collective Decision-Making

Humans flexibly integrate social information despite interindividual differences in reward

A multi-agent reinforcement learning model of reputation and cooperation in human groups

Collective action within an environment of unknown unknowns: Experiences with the port of Mars Game

Flocking in complex environments -- attention trade-offs in collective information processing

Using deep reinforcement learning to promote sustainable human behaviour on a common pool resource problem

Intrachoice Dynamics Shape Social Decisions

Collective intelligence: aggregation of information from neighbors in a guessing game

Social Learning in a Human Society: An Experimental Study

Predicting Human Behavior Toward Members of Different Social Groups.