摘要：在科学、医学、法律和金融等领域解决复杂任务，往往需要整合分散在规模远超模型上下文限制的海量异构来源中的、相互依赖的信息。现有方法通过将信息组织为更易于处理、可供模型推理的表示来应对这一挑战，例如图谱、文本记忆和检索集合。这些表示决定了下游推理能够实现什么，并最终决定推理是否成功；然而，其设计和构建在很大程度上仍具有临时性。在本研究中，我们借鉴相关性实现（relevance realization）的认知理论，提出了用于设计能够构建有效超大规模上下文表示的 AI 系统的具体原则。我们分析现有方法，并展示其成功与失败如何对应于它们与这些原则的一致程度；同时，我们推出 R3Con，一个旨在更系统地将这些原则付诸实践的工具框架。我们在两个近期的大型文档语料推理基准上，将 R3Con 与九种最先进的基线进行了评估。在这些基准上，R3Con 相比最强基线分别显著提升了 $20$ 和 $8.4$ 个百分点。它还使较小模型能够超过大得多的模型：采用 R3Con 的 4B 和 9B 模型超过了所有被评估的 35B 基线；而采用 R3Con 的 35B-A3B 模型，则以低 $3.7\times$ 的成本超过了搭载 Claude-Sonnet-5 的 Claude Code。我们的结果表明，遵循这一原则性方法构建的上下文表示可以降低对模型规模的依赖，指向一种由更小模型驱动、但具备前沿级性能的 AI 系统未来。我们的代码见此 HTTPS 链接

Abstract:Solving complex tasks in domains such as science, medicine, law, and finance often requires assembling interdependent information scattered across vast, heterogeneous sources far beyond model context limits. Existing approaches tackle this challenge by organizing information into more manageable representations over which models can reason, such as graphs, textual memories, and retrieval collections. These representations dictate what downstream reasoning is possible and, ultimately, whether it succeeds; yet their design and construction remain largely ad hoc. In this work, drawing on the cognitive theory of relevance realization, we propose concrete principles for designing AI systems that construct effective representations of very large contexts. We analyze existing approaches and show how their successes and failures map onto their alignment with these principles, and introduce R3Con, a harness designed to operationalize the principles more systematically. We evaluate R3Con against nine state-of-the-art baselines on two recent benchmarks of reasoning over large document corpora. On these benchmarks, R3Con substantially outperforms the strongest baseline, by $20$ and $8.4$ percentage points. It also enables smaller models to outperform much larger ones: R3Con with 4B and 9B models outperforms all evaluated 35B baselines, while R3Con with a 35B-A3B model outperforms Claude Code with Claude-Sonnet-5 at $3.7\times$ lower cost. Our results show that context representations following our principled approach can reduce reliance on model scale, pointing toward a future of AI systems with frontier-level performance powered by smaller models. Our code is available at this https URL