多阶段规则链框架用于组合与可解释的认知推理
A Multi-Stage Rule-Chaining Framework for Compositional and Interpretable Cognitive Reasoning
浏览论文内容
中文总结 AI 辅助
提出多阶段规则链框架，整合三个求解器在符号、结构、概念层面进行组合推理，在ARC基准上实现超95%准确率，提升可解释性与泛化能力。
中文摘要 AI 辅助
抽象与推理语料库（ARC）对认知泛化能力进行基准测试，即从有限示例中推断并应用抽象规则的能力。本文提出了一种多阶段规则链框架，在符号、结构和概念层面执行组合推理。该框架整合了三个互补的求解器：（1）确定性规则发现模块，通过几何、颜色和基于对象的分析诱导原子变换；（2）模式组合引擎，通过块合并、重复和空间启发式重构输出；（3）结构抽象层，推断网格间的层次和嵌套关系。这些求解器在渐进式回退层级中顺序运行，每个阶段重用先前的推理轨迹以增强可解释性和泛化能力。训练在1000个任务中通过了995个，进一步在120个任务中评估了105个，并解决了240个ARC-AGI-2任务中的230个测试任务。该系统在确定性、组合和抽象类别中实现了强覆盖，总体准确率超过95%。所提出的架构弥合了符号推理与模式合成之间的鸿沟，为认知泛化提供了可解释的洞见。结果表明，规则链和层次组合可以推动机器推理走向透明、与人类对齐的抽象，而无需依赖任务特定的调优。
英文摘要
The Abstraction and Reasoning Corpus (ARC) benchmarks cognitive generalization, the ability to infer and apply abstract rules from limited examples. This paper presents a multi-stage rule-chaining framework that performs compositional reasoning across symbolic, structural, and conceptual levels. The framework integrates three complementary solvers: (1) a deterministic rule discovery module that induces atomic transformations through geometric, color, and object-based analysis; (2) a pattern-composition engine that reconstructs outputs via block merging, repetition, and spatial heuristics; and (3) a structural abstraction layer that infers hierarchical and nested relationships across grids. These solvers operate sequentially within a progressive fallback hierarchy, where each stage reuses prior reasoning traces to enhance interpretability and generalization. Training passed for 995 tasks out of 1000, further evaluated on 105 tasks out of 120 and solved 230 test tasks out of 240 ARC-AGI-2 tasks. The system achieved strong coverage across deterministic, compositional, and abstract categories, demonstrating an overall accuracy exceeding 95 percent. The proposed architecture bridges symbolic reasoning and pattern synthesis, providing interpretable insight into cognitive generalization. The results suggest that rule chaining and hierarchical composition can advance machine reasoning toward transparent, human-aligned abstraction without relying on task-specific tuning.
发表机构
- Indian Institute of Technology Kharagpur（印度理工学院卡拉格普尔分校）
机构由 AI 辅助整理，请以论文原文为准。
