HP-GSBS：利用可复用区间统计加速贪心状态边界搜索
HP-GSBS: Accelerating greedy state boundary search with reusable interval statistics
arXiv:2610.25898v2 公告类型：replace-cross 摘要：贪婪状态边界搜索（Greedy State Boundary Search，GSBS）用于识别连续记录中相对稳定的神经活动模式之间的转换。重复的模板重建与相关性评估使得跨记录和跨参数设置的分析在计算上代价高昂。我们提出 HP-GSBS，一个开源实现，它将匹配目标表达为可加的区间贡献，并由可复用的相关统计量计算 t-distance。流式候选缩减、顺序局部细化与活动块复用，与 CPU、PyTorch CUDA 和 Triton 执行路径相辅相成。在七个受控工作负载（每个工作负载有两种输入实现）上，同输入加速比的中位数为 154 至 1,730。全部 168 次优化执行均复现了完整的参考边界历史与所选状态数，t-distance 的最大绝对差为 $1.88\times10^{-11}$。四个公开提供的 BBBD EEG 分析矩阵提供了互补的跨后端验证。一项自然主义 fMRI 应用对一组群体响应中的 4,230 个角回（angular-gyrus）体素以及 16 名参与者进行了分割；17 次 CPU 拟合总计耗时 213.10 秒。平均成对边界对齐程度超过循环移位基线，展示了所保存分割结果在队列层面的用途。带版本的结果、记录级并行以及可选的低内存执行支持了多样的分析规模。HP-GSBS 将针对重复计算的显式数学重构与高效且可复现的神经状态分析连接起来。
把 GSBS 的重复模板重建改写为可复用的区间相关统计，为跨记录、跨参数的神经状态分段提供开源加速实现。
arXiv:2610.25898v2 Announce Type: replace-cross Abstract: Greedy State Boundary Search (GSBS) identifies transitions between relatively stable neural activity patterns in continuous recordings. Repeated template reconstruction and correlation evaluation make analyses across recordings and parameter settings computationally expensive. We present HP-GSBS, an open-source implementation that expresses the matching objective as additive interval contributions and computes t-distance from reusable correlation statistics. Streaming candidate reduction, sequential local refinement, and active-block reuse complement CPU, PyTorch CUDA, and Triton execution paths. Across seven controlled workloads with two input realizations each, median same-input speedups range from 154 to 1,730. All 168 optimized executions reproduce complete reference boundary histories and selected state counts, with a maximum absolute t-distance difference of $1.88\times10^{-11}$. Four publicly supplied BBBD EEG analysis matrices provide complementary cross-backend validation. A naturalistic fMRI application segments 4,230 angular-gyrus voxels in a group response and 16 participants; the 17 CPU fits total 213.10 seconds. Mean pairwise boundary alignment exceeds a circular-shift baseline, illustrating cohort-level use of the saved segmentations. Versioned results, recording-level parallelism, and optional low-memory execution support varied analysis scales. HP-GSBS connects an explicit mathematical reformulation of repeated computation to efficient and reproducible neural-state analysis.
中文翻译还没有生成，先显示原文。
来源：arXiv q-bio.NC 预印本 · arxiv.org