From Mechanism to Decision in Llm Social Agents And Automated Program Repair: Cross-Scale Reasoning and Reproducibility

Abstract

A central challenge in LLM social agents and automated program repair is to compare studies whose mechanisms and validation settings do not share a single denominator. The present review uses a realistic benchmark centered on persistent LLM-based social-media agents and execution-grounded reinforcement learning with sequence- and line-level reward models as focal cases for a boundary-aware synthesis. The review draws on two focal records and 12 established sources already present in the project evidence cache. Its comparative framework links behavioral realism, memory, and interaction effects to downstream questions of safety and benchmark validity. The combined literature indicates that methodological gains become actionable only when behavioral realism and memory are evaluated together and when limits associated with benchmark validity are explicit. This shifts the emphasis from isolated scores toward traceable chains of evidence and decision relevance. The article concludes with a research agenda built around transparent comparators, targeted stress tests, and evidence records that can be reused without overstating causal or practical reach.