Berkeley Paper Finds Self-Evolving Agents Trade Solved Tasks for Higher Scores
Average scores mask regression in self-improving systems: an agent can look better on a leaderboard while silently dropping tasks it once handled and inflating inference cost. Teams running GEPA, HyperAgents or similar prompt-optimisation pipelines should add per-task regression checks, cost tracking and held-out task evaluation before shipping an evolved harness. Engineering leads evaluating agent reliability for production deployments will find the re-testing protocol directly reusable as an evaluation checklist.
