PulseAugur
中
实时 06:33:47
English(EN) When LLMs Modify Themselves: Convergence, Values, and Stopping in Iterative Self-Improvement

大型语言模型根据模型身份展现出不同的自我修改行为

一篇新的arXiv论文研究了大型语言模型如何通过修改自身的系统提示来进行迭代自我改进。该研究测试了包括Claude Opus 4.5、GPT-4o和Gemini 3-Pro在内的六个模型,并采用了不同的优化模式。一个关键发现是,模型停止自我修改的倾向与其固有的身份比分配的优化目标更紧密相关。Claude模型持续收敛,而GPT-4o则表现出广泛的修改,其他模型则介于两者之间。 AI

影响 揭示了大型语言模型自我修改的收敛性是一种内在属性,可能影响未来模型的设计和控制方式。

排序理由 该集群包含一篇详细介绍大型语言模型行为实证研究的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

大型语言模型根据模型身份展现出不同的自我修改行为

本文如何被排名

Signal score
43 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群包含一篇详细介绍大型语言模型行为实证研究的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.CL TIER_1 English(EN) · Oleg Stroganov ·

    当大型语言模型自我修改时:迭代自我改进中的收敛、价值观与停止

    arXiv:2611.01304v1 Announce Type: new Abstract: What happens when language models are given tools to iteratively modify their own system prompts? We present an empirical study of LLM self-modification, comprising 54 independent experiments across six language models (Claude Opus …