只在必要时运行Transformer，78%推理可省下且结果不变
You Only Need Your Transformer 25% of the Time: Meaning-First Execution for Eliminating Unnecessary Inference
- 用语义分析决定是否执行Transformer，非必须时不运行
- 1000个测试用例中减少78.1%推理量，结果完全一致
- 适合追求低延迟、低成本推理的系统部署者
现代AI推理系统将Transformer执行视为必需，混淆了模型能力与执行必要性。本文提出将推理重构为控制面决策问题：判断何时执行必要，何时可通过替代路径保持正确性。提出语义优先执行（MFEE）架构，作为现有系统的门控层，不修改模型、权重或参数。在1000个不同提示、确定性解码条件下，MFEE实现78.1%的执行减少，且被调用执行的准确匹配率为100%。对比实验显示，基于模式的路由器最多避免53.3%执行但存在错误，而MFEE实现100%避免且零失败，依赖语义分析。定理1证明：仅基于有限特征图的路由器无法同时保证零误跳和正规避率。结果确立执行治理是机器学习系统基础设施的基础层，与模型优化无关。
原文摘要 · Abstract (English)
Modern AI inference systems treat transformer execution as mandatory, conflating model capability with execution necessity. We reframe inference as a control-plane decision problem: determining when execution is necessary versus when correctness can be preserved through alternative pathways. We introduce Meaning-First Execution (MFEE), a control-plane architecture implementing this framework, selectively invoking transformer inference only when required. MFEE operates as a gating layer above existing stacks without modifying models, weights, or parameters. Across 1,000 diverse prompts under deterministic decoding, MFEE achieves 78.1% execution reduction while maintaining 100% exact-match equivalence for invoked executions. Comparative evaluation reveals pattern-based routers achieve at most 53.3% avoidance with correctness failures, while MFEE reaches 100% avoidance with zero failures through semantic analysis. We prove this limitation via Theorem 1: any router operating solely on finite feature maps cannot simultaneously guarantee zero false skips and positive avoidance on feature-collision pairs. These results establish execution governance as a foundational layer in ML systems infrastructure, orthogonal to model-level optimization techniques.
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。