Llm Social Agents And Automated Program Repair beyond Nominal Performance: Mechanisms, Uncertainty, and Deployment

Abstract

The literature on LLM social agents and automated program repair contains a recurring tension between methodological novelty and evidential comparability. By reading a realistic benchmark centered on persistent LLM-based social-media agents alongside execution-grounded reinforcement learning with sequence- and line-level reward models, this article clarifies the conditions under which their conclusions can support a common research argument. 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 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. On this basis, the review proposes an auditable pathway from focal mechanism to application claim, with explicit checkpoints for calibration, external validity, and responsible interpretation.