通过节点级计算成本编码，提升神经网络性能预测精度
Node-wise Feature Encoding for Neural Performance Prediction
- 在图注意力网络中显式编码每个节点的计算量、参数和内存
- 在延迟与能耗预测上均达当前最优，跨领域也表现稳定
- 适用于各类性能预测模型，添加成本极低
随着神经网络在资源受限的边缘设备上部署日益增多，准确预测延迟与能耗对高效神经架构搜索至关重要。现有基于图神经网络和Transformer的预测器虽表现优异，但普遍忽略节点级计算成本，难以建模关键性能操作。为此，本文提出FeatureFormer，一种在门控图注意力架构中显式融入各节点FLOPs、参数量及内存代理的神经性能预测器。同时，我们构建了NNEQ——一个大规模能耗数据集，支持延迟与能耗的统一评估。大量实验表明，FeatureFormer在两项指标上均达到当前最优，包括挑战性的跨领域场景。最后，我们验证该编码方法具有广泛适用性，可显著提升现有预测器性能，且开销极小。
原文摘要 · Abstract (English)
As neural networks are increasingly deployed on resource constrained edge devices, accurate prediction of latency and energy is critical for efficient neural architecture search. Existing GNN and transformer based predictors achieve strong results but largely ignore node-level computational cost, limiting their ability to model performance critical operations. To address this, we introduce FeatureFormer, a neural performance predictor that incorporates explicit node-wise encodings of FLOPs, parameter counts, and memory proxies within a gated graph attention architecture. We also present NNEQ, a new large-scale energy consumption dataset that enables unified evaluation of latency and energy prediction. Extensive experiments demonstrate that FeatureFormer achieves state-of-the-art performance across both metrics, including challenging out-of-domain settings. Finally, we show that the proposed encoding is broadly applicable and consistently improves existing predictors with negligible overhead.
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。