AI RESEARCH
When Does LLM Self-Correction Help? A Control-Theoretic Markov Diagnostic and Verify-First Intervention
arXiv CS.AI
•
ArXi:2604.22273v1 Announce Type: new Iterative self-correction is widely used in agentic LLM systems, but when repeated refinement helps versus hurts remains unclear. We frame self-correction as a cybernetic feedback loop in which the same language model serves as both controller and plant, and use a two-state Marko model over {Correct, Incorrect} to operationalize a simple deployment diagnostic: iterate only when ECR/EIR > Acc/(1 - Acc). In this view, EIR functions as a stability margin and prompting functions as lightweight controller design.