low-latency-inference

作者将 Gevva0 升级为级联决策引擎（静态规则 → 吸收马尔可夫链 → 快速路径 logits → 有界 CoT），并诚实地发布了对先前"100% GTO"说法的更正：修复提示泄漏 bug 后，零样本 GTO 一致率实为约 15%-29%，但延迟仍保持约 2ms。

SpikingVLA introduces an ANN-to-SNN conversion framework using Dendritic Integrate-and-Fire neurons and asynchronous execution for energy-efficient, low-latency spiking VLA inference, improving navigation success and reducing first-action latency by 11.2x over existing spiking VLA methods.