LATIDIA · Investigación
Inconsistencia de la memoria de múltiples agentes: dónde desaparecen los conflictos fácticos en los agentes LLM
arXiv:2611.22245v1 Tipo de anuncio: nuevo Resumen: La delegación multiactores permite a los sistemas de LLM distribuir la recopilación de evidencias entre trabajadores con contextos distintos. Esto plantea un problema de coherencia: los datos recopilados por los trabajadores
WhatsApp ↗Telegram ↗
La noticia
arXiv:2611.22245v1 Announce Type: new Abstract: Multi-agent delegation lets LLM systems distribute evidence collection across workers with separate contexts. This creates a consistency challenge: facts retrieved by workers may be missing from the evidence available to the lead for a joint conclusion. We introduce a benchmark and evidence-flow profiler for studying multi-agent memory inconsistency through controlled factual conflicts in Harvey LAB document-review tasks. Pooled across three models with comparable generation spending, requiring delegation increases retrieval of both conflicting claims from 50.2% to 68.7% relative to single-agent execution, yet conflict detection falls from 16.2% to 13.2%. Within this setting, workers report both claims for only 35.2% of retrieved pairs, and pooled matched