研究人员开发了两个独立的系统，分别命名为 Coral 和 CoRAL。Coral 是一个自适应系统，旨在跨异构云 GPU 高效地服务多个大语言模型，目标是优化资源分配并将服务成本降低高达 2.79 倍。另一方面，CoRAL 是一个用于机器人操作的框架，它使用大语言模型进行自适应控制，通过将高级推理与低级控制分离来实现零样本规划，并将接触丰富的场景下的成功率提高了 50% 以上。 AI
影响 引入了优化大语言模型服务成本和通过大语言模型集成增强机器人操作能力的新颖方法。
排序理由 提出了两篇不同的研究论文，一篇关于大语言模型服务基础设施，另一篇关于基于大语言模型的机器人控制。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2605.04357v1 Announce Type: cross Abstract: The usage of large language models (LLMs) has grown increasingly fragmented, with no single model dominating. Meanwhile, cloud providers offer a wide range of mid-tier and older-generation GPUs that enjoy better availability and d…
arXiv:2605.02600v1 Announce Type: cross Abstract: While Large Language Models (LLMs) and Vision-Language Models (VLMs) demonstrate remarkable capabilities in high-level reasoning and semantic understanding, applying them directly to contact-rich manipulation remains a challenge d…
The usage of large language models (LLMs) has grown increasingly fragmented, with no single model dominating. Meanwhile, cloud providers offer a wide range of mid-tier and older-generation GPUs that enjoy better availability and deliver comparable performance per dollar to top-ti…
While Large Language Models (LLMs) and Vision-Language Models (VLMs) demonstrate remarkable capabilities in high-level reasoning and semantic understanding, applying them directly to contact-rich manipulation remains a challenge due to their lack of explicit physical grounding an…