Llm Social Agents And Multimodal Executive Agents under Changing Conditions: From Local Mechanisms to System Decisions

Abstract

This review examines a shared methodological problem in LLM social agents and multimodal executive agents: how evidence from a realistic benchmark centered on persistent LLM-based social-media agents can be placed in analytical dialogue with a paired text-only and multimodal benchmark for constrained executive decision tasks without erasing differences in scale, assumptions, or intended use. Two target papers are triangulated against 12 locally validated publications. The comparison follows behavioral realism, memory, interaction effects, safety, benchmark validity and deliberately separates mechanistic interpretation from performance ranking, because the latter can conceal incompatible experimental or operational conditions. The synthesis shows that behavioral realism cannot be interpreted independently of memory, while interaction effects determines whether an apparent improvement remains meaningful outside the original setting. The strongest claims are therefore those that expose sensitivity, failure conditions, and residual uncertainty. The resulting framework supports reproducible comparison while preserving differences between study designs, and it identifies concrete points at which transfer claims should be narrowed or retested.