确保AI推理确定性是可信AI的基础，否则验证将失效。
On the Foundations of Trustworthy Artificial Intelligence
- 提出确定性假设，用信任熵量化非确定性代价。
- 实测82次跨架构推理零哈希差异，验证失败概率趋近于零。
- 适合关注AI可验证性、安全性和跨平台一致性的研究者。
我们证明，平台确定性推理是可信人工智能的必要且充分条件。将其形式化为确定性假说，并引入信任熵以量化非确定性的成本，精确证明验证失败概率等于1 - 2^{-H_T}。我们证明了确定性-验证坍缩：在确定性下，验证仅需常数级哈希比对；若无确定性，则验证者面临难以处理的成员资格问题。IEEE 754浮点运算本质上违背确定性要求。我们通过构建纯整数推理引擎，在ARM与x86架构间实现逐位相同输出。在高达67亿参数模型上进行的82次跨架构测试中，未出现任何哈希不匹配。四个地理分布节点产生完全一致输出，经由356笔链上证明交易验证。所有关键可信属性（公平性、鲁棒性、隐私、安全、对齐）均以平台确定性为前提。该系统由99,000行Rust代码构成，部署于三大洲，表明人工智能可信性本质是算术问题。
原文摘要 · Abstract (English)
We prove that platform-deterministic inference is necessary and sufficient for trustworthy AI. We formalize this as the Determinism Thesis and introduce trust entropy to quantify the cost of non-determinism, proving that verification failure probability equals 1 - 2^{-H_T} exactly. We prove a Determinism-Verification Collapse: verification under determinism requires O(1) hash comparison; without it, the verifier faces an intractable membership problem. IEEE 754 floating-point arithmetic fundamentally violates the determinism requirement. We resolve this by constructing a pure integer inference engine that achieves bitwise identical output across ARM and x86. In 82 cross-architecture tests on models up to 6.7B parameters, we observe zero hash mismatches. Four geographically distributed nodes produce identical outputs, verified by 356 on-chain attestation transactions. Every major trust property of AI systems (fairness, robustness, privacy, safety, alignment) presupposes platform determinism. Our system, 99,000 lines of Rust deployed across three continents, establishes that AI trust is a question of arithmetic.
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。