Tag
The paper introduces MILP-Evo, a closed-loop framework that uses LLM-guided program evolution to automatically design white-box MILP solver components (cut selectors and branching rules) by iteratively generating and evaluating candidate programs via end-to-end solver performance on MILP instances.
This paper challenges the prevailing claim that multi-agent systems outperform single-agent systems, demonstrating through systematic evaluation that automatically generated multi-agent architectures underperform Chain-of-Thought with Self-Consistency while being up to 10x more costly, and exposing architectural bloat in current automated design paradigms.
This paper proposes principled approaches for designing and optimizing practical agentic LLM systems, introducing a framework with pseudo-tools and fixed workflows to improve modularity, cost-efficiency, and accuracy across diverse tasks.