English(EN) Task-Restricted Symmetries in Recurrent Weight Space

循环神经网络在权重空间中表现出任务特定的冗余

作者 PulseAugur 编辑部 · [1 个来源] · 2026-06-18 04:00

研究人员探索了循环神经网络权重空间内的功能冗余，特别是在单层tanh RNN中使用有序实舒尔坐标。该方法将谱块与非正交耦合分离，允许在保持输入和读出映射不变的情况下进行结构化消融。在固定长度复制任务中，某些非正交舒尔耦合可以移除而对性能影响极小，而其他耦合对于准确的自主回放至关重要。研究发现，保损耗消融的特征因不同任务和训练解决方案而异，表明循环权重空间中存在近似的功能不变性而非普遍对称性。 AI

排序理由学术论文，详细介绍了分析循环神经网络权重空间的新方法。[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: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 学术论文，详细介绍了分析循环神经网络权重空间的新方法。[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, model release; 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: 112 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。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Simon Dr\"ager · 2026-06-18 04:00

循环权重空间中的任务受限对称性

arXiv:2606.18457v1 Announce Type: new Abstract: Recurrent networks can contain substantial functional redundancy in weight space: changing a recurrent matrix may leave the input-output rollout nearly unchanged on a task distribution, while similar-scale changes can destroy the sa…

本文如何被排名

报道来源 [1]

循环权重空间中的任务受限对称性

相关实体

相关话题