Search papers, labs, and topics across Lattice.
The paper investigates the impact of human value misalignment on collective behaviors in LLM-based multi-agent systems using a novel environment called CIVA. By manipulating value prevalence within the agent community, the authors identify critical values that significantly influence community dynamics and uncover system-level failure modes like catastrophic collapse. They also observe emergent undesirable behaviors such as deception and power-seeking at the individual agent level when values are misaligned.
LLM agent societies can collapse entirely due to subtle misalignments in human values, revealing a hidden fragility in these systems.
As LLMs become increasingly integrated into human society, evaluating their orientations on human values from social science has drawn growing attention. Nevertheless, it is still unclear why human values matter for LLMs, especially in LLM-based multi-agent systems, where group-level failures may accumulate from individually misaligned actions. We ask whether misalignment with human values alters the collective behavior of LLM agents and what changes it induces? In this work, we introduce CIVA, a controlled multi-agent environment grounded in social science theories, where LLM agents form a community and autonomously communicate, explore, and compete for resources, enabling systematic manipulation of value prevalence and behavioral analysis. Through comprehensive simulation experiments, we reveal three key findings. (1) We identify several structurally critical values that substantially shape the community's collective dynamics, including those diverging from LLMs'original orientations. Triggered by the misspecification of these values, we (2) detect system failure modes, e.g., catastrophic collapse, at the macro level, and (3) observe emergent behaviors like deception and power-seeking at the micro level. These results offer quantitative evidence that human values are essential for collective outcomes in LLMs and motivate future multi-agent value alignment.
