English(EN) The Internal Anatomy of Strategic Choice in Large Language Models

新研究分析大型语言模型的战略选择机制

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-10 04:00

研究人员调查了大型语言模型的内部决策过程，特别是检查了它们如何在博弈论场景中处理战略选择。通过记录模型在各种 $2\times2$ 游戏的单次博弈中的激活情况，研究发现包括 Qwen2.5 在内的密集模型和混合专家模型都能检测并响应激励。然而，模型在激励影响其选择的方式上表现出差异，其中一些模型表明，训练后修改可以在不显著改变可观察行为的情况下，改变从表示的激励到决策的计算路径。 AI

影响为了解大型语言模型如何处理战略决策提供了见解，可能有助于开发更复杂的 AI 代理。

排序理由分析大型语言模型行为的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

Qwen2.5

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

本文如何被排名

Signal score: 13 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 分析大型语言模型行为的学术论文。[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]

arXiv cs.AI TIER_1 English(EN) · Vin\'icius Ferraz, Leon Houf, Enrico Ferrea · 2026-09-10 04:00

大型语言模型中战略选择的内部解剖

arXiv:2609.07478v1 Announce Type: new Abstract: Large language models act as strategic agents and models of human choice, yet choosing like a strategic agent does not mean computing like one. We recorded activations from four open-weight models --- dense and mixture-of-experts, i…

本文如何被排名

报道来源 [1]

大型语言模型中战略选择的内部解剖

相关实体

相关话题