LATIDIA · Investigación

De las preguntas a las decisiones: evaluación comparativa de cómo los agentes de LLM adquieren información relevante para la decisión de los seres humanos

arXiv: 2611.13015v1Tipo de anuncio: nuevo Resumen: A los agentes de LLM se les pide cada vez más que hagan planes para las personas, como programar conductores o asignar enfermeras. Sin embargo, las solicitudes de las personas a menudo están incompletas y el soporte

La noticia

arXiv:2611.13015v1 Announce Type: new Abstract: LLM agents are increasingly asked to make plans for people, such as scheduling drivers or assigning nurses. However, people's requests are often incomplete, and the supporting information for decision-making may be outdated, requiring agents to ask for clarification. Existing benchmarks typically provide fully specified problems or rely on uncalibrated LLM-simulated users, leaving this ability insufficiently tested. We introduce Ask2Decide, a benchmark of 43 planning tasks that evaluates whether agents ask for missing information that matters for their decisions. An operator holds facts absent from the agent's request and records, and optimization solvers determine which facts change the optimal plan. We also include similarly worded facts