PulseAugur
中
实时 19:47:41
English(EN) Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention

NVIDIA 发布 Gated DeltaNet-2 以改进线性注意力

NVIDIA 推出了 Gated DeltaNet-2,这是一种新的线性注意力层,旨在改进循环神经网络中的内存编辑。该模型使用独立的通道门控机制,将擦除旧信息和写入新信息的过程分离开来,解决了先前 delta-rule 架构中的局限性。Gated DeltaNet-2 在 1000 亿 token 和 13 亿参数上进行了训练,在长上下文检索任务上表现优于 Mamba-2 和 KDA 等现有模型。 AI

影响 增强了循环模型中的长上下文处理能力,有望提高复杂语言任务的性能。

排序理由 该集群描述了一种新的模型架构及其在基准测试上的性能,该模型在 arXiv 论文中有详细介绍,并被科技新闻媒体报道。

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →

NVIDIA 发布 Gated DeltaNet-2 以改进线性注意力

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
Research
该集群描述了一种新的模型架构及其在基准测试上的性能,该模型在 arXiv 论文中有详细介绍,并被科技新闻媒体报道。
Source corroboration
4 independent sources
Strong cross-source corroboration — multiple independent publishers covered this within the clustering window.
Topics
model release, 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
140 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.
Coverage growth since scoring
+1 source(s) since last score
New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

完整方法见我们的编辑标准。

报道来源 [4]

  1. arXiv cs.AI TIER_1 English(EN) · Jan Kautz ·

    Gated DeltaNet-2: 解耦线性注意力中的擦除和写入

    Linear attention replaces the unbounded cache of softmax attention with a fixed-size recurrent state, reducing sequence mixing to linear time and decoding to constant memory. The hard part is not just what to forget, but how to edit this compressed memory without scrambling exist…

  2. MarkTechPost TIER_1 English(EN) · Asif Razzaq ·

    NVIDIA AI 发布 Gated DeltaNet-2:一种解耦 Delta 规则中擦除和写入的线性注意力层

    <p>Linear attention squeezes the unbounded KV cache into a fixed-size recurrent state, but editing that memory without scrambling existing associations is hard. Prior delta-rule models like Gated DeltaNet and KDA use one scalar gate to control both erasing old content and writing…

  3. Mastodon — fosstodon.org TIER_1 Polski(PL) · [email protected] ·

    NVIDIA发布Gated DeltaNet-2,一种新的线性注意力架构,通过独立的读写门大幅提高了精度

    NVIDIA zaprezentowała Gated DeltaNet-2, nową architekturę liniowej atencji, która dzięki niezależnym bramkom zapisu i usuwania danych drastycznie poprawia precyzję modeli AI w długich kontekstach. # si # ai # sztucznainteligencja # wiadomości # informacje # technologia https:// a…

  4. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] ·

    NVIDIA 发布 Gated DeltaNet-2,一种线性注意力层,通过独立的通道门控解耦了擦除旧内容与写入新内容。在 1

    NVIDIA has released Gated DeltaNet-2, a linear attention layer that decouples erasing old content from writing new content via separate channel-wise gates. At 1.3B parameters trained on 100B tokens, it outperforms Mamba-2, Gated DeltaNet and KDA on language modelling and long-con…