Explicit Trait Inference for Multi-Agent Coordination

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Abdurahman, Suhaib, Ishii, Etsuko, Margatina, Katerina, Bhargavi, Divya, Sunkara, Monica, Zhang, Yi
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913054589976576
author Abdurahman, Suhaib
Ishii, Etsuko
Margatina, Katerina
Bhargavi, Divya
Sunkara, Monica
Zhang, Yi
author_facet Abdurahman, Suhaib
Ishii, Etsuko
Margatina, Katerina
Bhargavi, Divya
Sunkara, Monica
Zhang, Yi
contents LLM-based multi-agent systems (MAS) show promise on complex tasks but remain prone to coordination failures such as goal drift, error cascades, and misaligned behaviors. We propose Explicit Trait Inference (ETI), a psychologically grounded method for improving coordination. ETI enables agents to infer and track partner characteristics along two established psychological dimensions--warmth (e.g., trust) and competence (e.g., skill)--from interaction histories to guide decisions. We evaluate ETI in controlled settings (economic games), where it reduces payoff loss by 45-77%, and in more realistic, complex multi-agent settings (MultiAgentBench), where it improves performance by 3-29% depending on the scenario and model, relative to a CoT baseline. Additional analysis shows that gains are closely linked to trait inference: ETI profiles predict agents' actions, and informative profiles drive improvements. These results highlight ETI as a lightweight and robust mechanism for improving coordination in diverse multi-agent settings, and provide the first systematic evidence that LLM agents can (i) reliably infer others' traits from interaction histories and (ii) leverage structured awareness of others' traits for coordination.
format Preprint
id arxiv_https___arxiv_org_abs_2604_19278
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Explicit Trait Inference for Multi-Agent Coordination
Abdurahman, Suhaib
Ishii, Etsuko
Margatina, Katerina
Bhargavi, Divya
Sunkara, Monica
Zhang, Yi
Artificial Intelligence
Multiagent Systems
LLM-based multi-agent systems (MAS) show promise on complex tasks but remain prone to coordination failures such as goal drift, error cascades, and misaligned behaviors. We propose Explicit Trait Inference (ETI), a psychologically grounded method for improving coordination. ETI enables agents to infer and track partner characteristics along two established psychological dimensions--warmth (e.g., trust) and competence (e.g., skill)--from interaction histories to guide decisions. We evaluate ETI in controlled settings (economic games), where it reduces payoff loss by 45-77%, and in more realistic, complex multi-agent settings (MultiAgentBench), where it improves performance by 3-29% depending on the scenario and model, relative to a CoT baseline. Additional analysis shows that gains are closely linked to trait inference: ETI profiles predict agents' actions, and informative profiles drive improvements. These results highlight ETI as a lightweight and robust mechanism for improving coordination in diverse multi-agent settings, and provide the first systematic evidence that LLM agents can (i) reliably infer others' traits from interaction histories and (ii) leverage structured awareness of others' traits for coordination.
title Explicit Trait Inference for Multi-Agent Coordination
topic Artificial Intelligence
Multiagent Systems
url https://arxiv.org/abs/2604.19278