让神经网络学会物理基本规律，实现从猜数据到懂原理的跨越。
From Phenomenological Fitting to Endogenous Deduction: A Paradigm Leap via Meta-Principle Physics Architecture
- 将物理三大元原理嵌入网络结构，构建可推演的物理认知框架。
- 在物理推理等任务上性能显著提升，逻辑与数学能力分别提高52%和2.18倍。
- 适合追求可解释性、物理常识与因果推理的AI研究者使用。
当前神经网络本质是现象拟合：通过海量参数和数据学习输入输出间的统计关联，却缺乏对物理现实基本规律的内在理解。本文提出从纯现象拟合到现象拟合与内生推导融合的范式跃迁。通过将物理元原理嵌入网络架构，构建元原理物理架构（MPPA）。具体地，MPPA嵌入连接性、守恒性、周期性三大核心元原理，分别由引力器（标准因果注意力实现连接性）、能量编码器（对数域能量追踪与延迟补偿实现守恒性）、周期性编码器（基于FFT的谱分析与延迟调制实现周期性）实现。三者通过可学习独立门控融合机制协同工作，形成‘局部关系连接-全局守恒约束-演化周期律’的完整物理认知框架。实验表明，MPPA在物理推理任务上从接近零提升至0.436（0.436 vs 0.000），数学任务提升2.18倍（0.330 vs 0.151），逻辑任务提升52%（0.456 vs 0.300），验证集困惑度降低3.69%（259.45 vs 269.40），仅增加11.8%参数量（242.40M vs 216.91M）。尤其在分布外物理场景中表现出强泛化能力，验证了该原理嵌入设计的鲁棒性与可解释性。本工作为具备物理常识、因果推理与数学严谨性的下一代AI奠定了理论基础与技术路径。
原文摘要 · Abstract (English)
The essence of current neural network architectures is phenomenological fitting: they learn input-output statistical correlations via massive parameters and data, yet lack intrinsic understanding of the fundamental principles governing physical reality. This paper proposes a paradigm leap from pure phenomenological fitting to the fusion of phenomenological fitting and endogenous deduction. By embedding physical meta-principles into neural network architecture, we construct the Meta-Principle Physics Architecture (MPPA). Specifically, MPPA embeds three core meta-principles - Connectivity, Conservation, Periodicity - into its architecture, implemented via three core components: the Gravitator realizes Connectivity via standard causal attention; the Energy Encoder implements Conservation via log-domain energy tracking and delayed compensation; the Periodicity Encoder fulfills Periodicity via FFT-based spectral analysis and delayed modulation. These components collaborate via a learnable independent gating fusion mechanism, forming a complete physical cognition framework of 'local relational connectivity - global conservation constraint - evolutionary periodic law'. Experiments show MPPA achieves significant improvements: physical reasoning (from near zero to 0.436, 0.436 vs 0.000), 2.18x mathematical task improvement (0.330 vs 0.151), 52% logical task gain (0.456 vs 0.300), and 3.69% lower validation perplexity (259.45 vs 269.40), with only 11.8% more parameters (242.40M vs 216.91M). Notably, MPPA shows strong generalization on out-of-distribution physical scenarios, proving the robustness and interpretability of this principle-embedded design. This work establishes a new theoretical foundation and technical path for next-generation AI with physical common sense, causal reasoning, and mathematical rigor.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。