English(EN) Is spectral predictability enough? Testing {\Omega}-matching for synthetic pretraining under distribution shift

光谱可预测性不足以在分布偏移下进行合成预训练

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-23 04:00

一篇新的研究论文探讨了光谱可预测性（$\Omega$）匹配在机器学习模型合成预训练中的有效性，特别是在分布偏移的情况下。虽然$\Omega$匹配的预训练在受控的合成到合成场景中显示出显著的改进，但在ETtm1等真实基准上的表现却微乎其微，甚至适得其反。研究发现，仅凭$\Omega$不足以预测跨不同模型架构和数据源的有利数据迁移，这表明需要额外的信号属性才能有效选择源。 AI

影响指出了当前合成数据选择方法的局限性，可能指导未来在更鲁棒的预训练策略方面的研究。

排序理由在arXiv上发表的研究论文，详细介绍了一种新的合成预训练方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

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: 5 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 在arXiv上发表的研究论文，详细介绍了一种新的合成预训练方法。[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, other; 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: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Edric Castel C. Hao, Juan Raphael S. Se\~na, Neon S. De Jesus · 2026-10-23 04:00

Is spectral predictability enough? Testing {\Omega}-matching for synthetic pretraining under distribution shift

arXiv:2610.30802v1 Announce Type: new Abstract: Synthetic pretraining under distribution shift requires deciding what synthetic data to use. A natural criterion is to match the spectral predictability $\Omega$ of the source to the target, but is $\Omega$ sufficient? We test this …

本文如何被排名

报道来源 [1]

Is spectral predictability enough? Testing {\Omega}-matching for synthetic pretraining under distribution shift

相关实体

相关话题