Valori：AI系统的确定性内存基座

Valori: A Deterministic Memory Substrate for AI Systems

December 25, 2025

作者: Varshith Gudur

cs.AI

摘要

现代人工智能系统依赖基于浮点运算存储和检索的向量嵌入。虽然这种设计能有效实现近似相似性搜索，但其本质引入了非确定性：即使模型、输入数据和代码完全相同，在不同硬件架构（如x86与ARM）上也会产生不同的内存状态和检索结果。这导致系统无法实现状态复现与安全部署，引发难以察觉的数据偏差，使得受监管领域的事后验证与审计追踪难以进行。我们提出Valori——一种确定性AI内存基座，通过定点运算（Q16.16格式）替代浮点内存操作，并将内存建模为可复现状态机。Valori能确保跨平台的比特级一致内存状态、快照及搜索结果。我们论证了非确定性在索引或检索操作之前就已产生，并展示Valori如何在内存边界实施确定性保障。研究结果表明，确定性内存是构建可信AI系统的必要基础组件。该参考实现已开源（项目地址：https://github.com/varshith-Git/Valori-Kernel，归档于https://zenodo.org/records/18022660）。

English

Modern AI systems rely on vector embeddings stored and searched using floating-point arithmetic. While effective for approximate similarity search, this design introduces fundamental non-determinism: identical models, inputs, and code can produce different memory states and retrieval results across hardware architectures (e.g., x86 vs. ARM). This prevents replayability and safe deployment, leading to silent data divergence that prevents post-hoc verification and compromises audit trails in regulated sectors. We present Valori, a deterministic AI memory substrate that replaces floating-point memory operations with fixed-point arithmetic (Q16.16) and models memory as a replayable state machine. Valori guarantees bit-identical memory states, snapshots, and search results across platforms. We demonstrate that non-determinism arises before indexing or retrieval and show how Valori enforces determinism at the memory boundary. Our results suggest that deterministic memory is a necessary primitive for trustworthy AI systems. The reference implementation is open-source and available at https://github.com/varshith-Git/Valori-Kernel (archived at https://zenodo.org/records/18022660).