研究人员推出了一种新颖的自我改进AI系统SIA，该系统可同时更新智能体的脚手架（工具、提示、逻辑）及其模型权重。这一方法在最新论文中进行了详细介绍，它结合了两个先前独立的研究方向：脚手架更新和测试时训练。通过整合这些方法，SIA旨在克服AI开发和改进中的人为瓶颈。在法律分类、GPU内核优化和RNA去噪任务上的评估表明，与仅更新脚手架或权重的现有方法相比，性能有了显著提升。 AI
影响 这种方法可以通过减少人为干预来加速AI开发，可能带来更强大、更具适应性的AI系统。
排序理由 该集群描述了一篇关于新颖AI代理架构及其性能评估的最新研究论文。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
arXiv:2605.27276v1 Announce Type: new Abstract: Humans are the bottleneck in building and improving AI. Both the models and the agents that wrap them are written, tuned, and corrected by people. The long-horizon goal of an AI that can figure out how to improve itself remains open…
Humans are the bottleneck in building and improving AI. Both the models and the agents that wrap them are written, tuned, and corrected by people. The long-horizon goal of an AI that can figure out how to improve itself remains open. Two largely disjoint research lines attack thi…
<p>Hexo Labs released SIA, an open-source self-improving loop, under an MIT license. A Feedback-Agent reads each run's trajectory, then either rewrites the scaffold or triggers a LoRA weight update on gpt-oss-120b. Combining both levers beat scaffold-only iteration on LawBench, T…