English(EN) Light Origins Open-Sources LightNav-0 Generalist Navigation Brain

Light Origins 开源 LightNav-0 通用导航模型

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-17 08:05

Light Origins 发布了 LightNav-0，一个基于 Qwen3-VL-4B 的通用导航模型。该模型使用 Real2Sim2Real 方法，在超过 2000 个场景和 4000 小时的 VLA 数据上进行了训练。LightNav-0 表现强劲，在 10 项单目导航基准测试中均处于领先地位，并具备零样本迁移能力。 AI

影响此次发布为从事机器人导航和具身人工智能研究的开发者提供了一个新工具。

排序理由专业 AI 模型开源发布。[lever_c_demoted from research: ic=1 ai=1.0]

在 Pandaily 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: 29 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 专业 AI 模型开源发布。[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: model release, product; 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: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

Pandaily TIER_1 English(EN) · [email protected] (Pandaily) · 2026-09-17 08:05

Light Origins 开源 LightNav-0 通用导航大脑

Light Origins open-sourced LightNav-0, a Qwen3-VL-4B navigation model trained via Real2Sim2Real on 2,000+ scenes and 4,000+ hours of VLA data, leading 10 monocular navigation benches with zero-shot body transfer.

本文如何被排名

报道来源 [1]

Light Origins 开源 LightNav-0 通用导航大脑

相关实体

相关话题