你只需要在25%的时间使用你的Transformer:以意义优先执行以消除不必要的推理
You Only Need Your Transformer 25% of the Time: Meaning-First Execution for Eliminating Unnecessary Inference
AI总结:
本文提出意义优先执行框架,通过控制平面决策减少不必要的Transformer推理,实现78.1%的执行减少和100%的正确性保障。
摘要:
现代AI推理系统将Transformer执行视为必需的,将模型能力与执行必要性混淆。我们重新将推理视为一个控制平面决策问题:确定何时执行是必要的,何时可以通过替代路径保持正确性。我们引入了意义优先执行(MFEE),一种实现此框架的控制平面架构,仅在需要时调用Transformer推理。MFEE在现有堆栈之上作为门控层运行,不修改模型、权重或参数。在1000个多样化的提示下,通过确定性解码,MFEE实现了78.1%的执行减少,同时保持调用执行的100%精确匹配等价性。比较评估显示,基于模式的路由器最多只能实现53.3%的规避,但存在正确性失败;而MFEE通过语义分析实现了100%的规避,无任何错误。我们通过定理1证明了这一限制:任何仅基于有限特征图的路由器无法同时保证零假跳过和在特征冲突对上的正规避。这些结果确立了执行治理作为机器学习系统基础设施的基础层,与模型级优化技术正交。
英文摘要:
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.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。