Tech Meridian ← LIVE FEED
PROMY MERIDIAN RU

RESEARCH · RESEARCH · #3970

GRAFT: execution-guided genetic recombination for automated agent design (arXiv:2611.23509v1)

The paper introduces GRAFT (Execution Guided Genetic Recombination for Agent Design), an evolutionary search method that uses per-parent execution evidence to guide semantic crossover and integrated mutation when designing LLM-based agent architectures. Evaluated across six reasoning and programming tasks with open-weight optimizer/executor models, GRAFT achieves a mean test accuracy of 82.0%, outperforming the strongest baseline by 9.8 percentage points without requiring extra task executions.

KEY POINTS

  1. The paper introduces GRAFT (Execution Guided Genetic Recombination for Agent Design), an evolutionary search method that uses per-parent execution evidence to guide semantic crossover and integrated mutation when designing LLM-based agent architectures.
  2. Evaluated across six reasoning and programming tasks with open-weight optimizer/executor models, GRAFT achieves a mean test accuracy of 82.0%, outperforming the strongest baseline by 9.8 percentage points without requiring extra task executions.
  3. Shows that reuse of instance-level execution evidence can substantially improve evolutionary search for LLM-based agent architectures, offering a practical way to boost agent performance without extra evaluations.

WHY IT MATTERS

Shows that reuse of instance-level execution evidence can substantially improve evolutionary search for LLM-based agent architectures, offering a practical way to boost agent performance without extra evaluations.

SOURCES & TIMELINE

1