English(EN) Operator-Aware Mixed-Precision Tolerance Calibration for Tensor Kernels

新方法提升AI计算张量核的错误检测能力

作者 PulseAugur 编辑部 · [1 个来源] · 2026-07-21 04:00

研究人员开发了一种新的方法来校准张量核中的混合精度容差，张量核对于机器学习计算至关重要。这种面向算子的方法分析了云GPU运行中的逐元素误差分布，以建立更严格、有实证依据的绝对容差。当应用于LLM风格的错误变体时，该方法将错误检测召回率显著提高了9个百分点以上，识别出200多个额外错误。 AI

影响通过提高张量核正确性测试的准确性，提升了AI模型训练和推理的可靠性。

排序理由该条目是一篇学术论文，详细介绍了一种校准张量核的新方法。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该条目是一篇学术论文，详细介绍了一种校准张量核的新方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, infra; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 83 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

arXiv cs.LG TIER_1 English(EN) · Dipankar Sarkar · 2026-07-21 04:00

面向张量核的算子感知混合精度容差校准

arXiv:2607.16228v1 Announce Type: new Abstract: Most tensor-kernel correctness tests go through a fixed-shape all close-style check with hand-picked absolute and relative tolerances. The thresholds are copied across the corpus and rarely revisited. We mine the element-wise error …