研究人员推出了ASPIRE，这是一个新的基准测试，旨在测试大型语言模型（LLM）在给定模糊的自然语言目标而非明确的任务和指标时的自我进化能力。ASPIRE基准要求代理解释这些目标，识别学习差距，选择合适的数据和更新方法，并确定自己的进展。实验表明，虽然目前的代理可以驾驭训练和编辑循环，但实现稳定的权重级收益具有挑战性，并且它们的性能通常不如Qwen-Agent等工程化参考。 AI
影响 该基准测试可能会推动LLM朝着更自主的学习和目标解释方向发展，从而可能产生更具适应性的AI系统。
排序理由 该集群描述了在arXiv上发布的一个用于LLM自我进化的新基准测试。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2608.31111v1 Announce Type: new Abstract: Many important forms of human learning begin with a vague goal, such as "become a better physicist" or "improve at research." Learners must interpret the goal, identify capability gaps, decide how to learn, and determine whether the…
Many important forms of human learning begin with a vague goal, such as "become a better physicist" or "improve at research." Learners must interpret the goal, identify capability gaps, decide how to learn, and determine whether they have actually improved. In contrast, existing …
ASPIRE introduces a benchmark for self-evolving LLM agents from vague natural-language goals, revealing challenges in goal interpretation, data selection, and stable weight-level improvement.