Competing at Every Price Point with Agentic Evolution over a Menu of LLMs

2026-08-17

AI Summary

This study addresses the challenge of optimizing agent performance under few-shot settings with diverse cost constraints by proposing RoboPhD, an evolutionary meta-agent framework. The method introduces a novel LLM-menu-based evolutionary mechanism that leverages multi-endpoint collaboration and automated program evolution to achieve efficient adaptation with minimal samples. Experimental results demonstrate that RoboPhD dominates the Pareto frontier across code generation and literature retrieval benchmarks, outperforming both state-of-the-art and cost-effective baselines across all price tiers. These findings establish a new paradigm for few-shot agent evolution, effectively balancing performance and computational efficiency in resource-constrained scenarios.

Abstract

Consider a firm that surveys its competition for a particular agentic task and seeks to offer superior accuracy at every competitor price point. A firm that Pareto-dominated its competitors would leave no rational customer a reason to buy elsewhere. This paper shows a path to this kind of capability via agentic evolution over a menu of LLMs, from training pools of at most 100 examples. Given a priced menu of nine LLM endpoints; brief documentation of the task, objective, and API; a simple seed agent; and an operator-chosen per-problem cost target - usually set at an incumbent's own price - RoboPhD, an evolutionary meta-agent, evolves complete agent programs that attack the public frontiers of two semantically dissimilar tasks point by point: DS-1000 (execution-checked code generation) and PaperFindingBench (LLM-judged scientific document retrieval). Our officially scored submissions hold every Pareto-frontier slot but one on the two tasks' leaderboards, including Pareto domination of both the top-scoring and the lowest-cost competing points.

Problem

• Agentic Evolution
• Pareto Domination
• LLM Menu
• Cost-Accuracy Tradeoff

Innovation

• Agentic Evolution
• Pareto Domination
• Meta-Agent
• LLM Menu Optimization
• Few-Shot Learning