Current context augmentation methods, such as retrieval-augmented generation, are essential for solving knowledge-intensive reasoning tasks. However, they typically adhere to a rigid, brute-force strategy that executes retrieval at every step. This indiscriminate approach not only incurs unnecessary computational costs but also degrades performance by saturating the context with irrelevant noise. To address these limitations, we introduce Agentic Context Evolution (ACE), a framework inspired by human metacognition that dynamically determines whether to seek new evidence or reason with existing knowledge. ACE employs a central orchestrator agent to make decisions strategically via majority voting. It aims to alternate between activating a retriever agent for external retrieval and a reasoner agent for internal analysis and refinement. By eliminating redundant retrieval steps, ACE maintains a concise and evolved context. Extensive experiments on challenging multi-hop QA benchmarks demonstrate that ACE significantly outperforms competitive baselines in accuracy while achieving efficient token consumption. Our work provides valuable insights into advancing context-evolved generation for complex, knowledge-intensive tasks.

翻译：当前基于上下文增强的方法，如检索增强生成，对于解决知识密集型推理任务至关重要。然而，这些方法通常遵循一种僵化的、蛮力式的策略，即在每一步都执行检索。这种不加区分的做法不仅会产生不必要的计算成本，还会因在上下文中充斥无关噪声而降低性能。为了应对这些局限性，我们引入了智能体上下文演化框架，该框架受人类元认知启发，能够动态决定是寻求新证据还是基于现有知识进行推理。ACE采用一个中央编排智能体，通过多数投票机制进行策略性决策。其目标是在激活检索智能体以进行外部检索，与激活推理智能体以进行内部分析和精炼之间交替进行。通过消除冗余的检索步骤，ACE得以维持一个简洁且不断演化的上下文。在具有挑战性的多跳问答基准上进行的大量实验表明，ACE在准确率上显著优于竞争性基线方法，同时实现了高效的令牌消耗。我们的工作为推进面向复杂、知识密集型任务的上下文演化生成提供了有价值的见解。