一篇新的研究论文探讨了语言模型代理如何从其“约束”（包括提示、工具和控制逻辑）中学习偏好。研究表明，策略的已学偏好与其训练约束相关，当约束改变时会影响其性能。通过分析“上限”（约束潜力）和“不匹配”（策略次优性），研究人员可以更好地预测模型在新环境中的适应能力。 AI
影响 这项研究可能有助于开发更强大的AI代理，使其能够更有效地适应不同的任务和环境。
排序理由 在arXiv上发表的研究论文，详细介绍了关于AI代理接口的发现。
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:2610.37255v1 Announce Type: cross Abstract: Language model agents are post-trained and deployed through a harness, the prompts, tools, and control logic that present a task to the model and execute its outputs. Harnesses are often revised and judged by how an already traine…
Language model agents are post-trained and deployed through a harness, the prompts, tools, and control logic that present a task to the model and execute its outputs. Harnesses are often revised and judged by how an already trained model performs under them, which assumes that wh…
Under KL-regularized control, a policy encodes its harness preferences, so deployed performance equals harness potential minus policy mismatch, and harness comparisons must separate these terms to predict training gains.