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
- 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.
- 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.