引入了一个名为AutoLab的新基准，用于评估前沿AI模型的长时域迭代优化能力。该基准包含四个领域的36个任务，要求智能体在时间预算内改进次优基线。对17个最先进模型的评估表明，坚持性和时间意识比初始性能对成功更重要，Anthropic的Claude Opus 4.6展示了强大的能力，而许多其他模型则在过早终止或进展甚微方面遇到困难。 AI
影响 强调了AI智能体在复杂、长期任务中发展坚持性和时间意识的必要性。
排序理由 该集群描述了一篇介绍AI研究基准的新学术论文。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2606.05080v1 Announce Type: new Abstract: Scientific and engineering progress is fundamentally a long-horizon iterative process: proposing changes, running experiments, measuring outcomes, and continuously refining artifacts. Yet existing benchmarks for frontier models prim…
Scientific and engineering progress is fundamentally a long-horizon iterative process: proposing changes, running experiments, measuring outcomes, and continuously refining artifacts. Yet existing benchmarks for frontier models primarily evaluate either single-turn responses or s…
Scientific and engineering progress is fundamentally a long-horizon iterative process: proposing changes, running experiments, measuring outcomes, and continuously refining artifacts. Yet existing benchmarks for frontier models primarily evaluate either single-turn responses or s…
AutoLab benchmark evaluates long-horizon iterative optimization capabilities of frontier models across diverse domains, revealing that persistent iteration and time awareness are more critical than initial performance quality.