arXiv 2607.014332026-07-03cs.AIcs.LG

CreativityNeuro: 引导语言模型权重以改善发散思维并减少模式崩溃

CreativityNeuro: Steering Language Model Weights to Improve Divergent Thinking and Reduce Mode Collapse

Center for Brain Science, Harvard University（哈佛大学脑科学中心）；CBS-NTT Program in Physics of Intelligence, Harvard University（哈佛大学CBS-NTT智能物理项目）；Prior Computers；AI Innovation Institute, Stony Brook University（石溪大学人工智能创新研究所）

Samuel Schapiro, Core Francisco Park, Felix Sosa, Lav R. Varshney

备注：: Accepted at ICML 2026 Workshop on Creativity & Generative AI

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出数据无关的对比权重引导方法CreativityNeuro，通过调整LLM权重增强发散思维，在多项创造力测试中提升原创性并减少模式崩溃，无需重新训练或微调。

摘要：

发散思维是创造力的关键方面，然而大型语言模型（LLMs）倾向于对开放式问题持续生成相似回答，这被称为人工蜂群效应。在此，我们介绍CreativityNeuro，一种通过对比权重引导增强LLM发散思维的无数据方法。我们在多项创造力评估中评估了该方法，并报告了几个主要发现。在发散联想任务（DAT）中，一种词汇空间创造力测试，CreativityNeuro将性能提升了多达14个人类百分位点。接下来，在关于替代用途测试（AUT）和任务任务的大规模人类评估（N=720）中，CreativityNeuro在原创性、惊喜度和创造力方面取得了显著改进，并迁移到更长形式和更开放的任务中。重要的是，我们发现，在所有三个任务中，CreativityNeuro明显减少了模式崩溃的度量。此外，激活引导在DAT上达到了与CreativityNeuro相当的性能，但未能迁移到AUT和任务任务，证明了权重空间引导在泛化到未见任务上的有效性。总之，CreativityNeuro无需行为数据、重新训练或基于梯度的微调即可改善发散思维并减少模式崩溃，为增强LLM在创造性领域的性能提供了一种直接的方法。

英文摘要：

Divergent thinking is a crucial aspect of creativity, yet large language models (LLMs) tend to consistently generate similar responses to open-ended questions, in what has been termed the artificial hivemind effect. Here, we introduce CreativityNeuro, a data-free method for enhancing divergent thinking in LLMs via contrastive weight steering. We evaluate our method across multiple creativity assessments and report several main findings. On the Divergent Association Task (DAT), a vocabulary-space creativity test, CreativityNeuro improves performance by up to 14 human percentile points. Next, in a large-scale human evaluation (N=720) on the Alternative Uses Test (AUT) and the Task Task, CreativityNeuro achieves significant improvements in originality, surprise, and creativity, transferring to longer-form and more open-ended tasks. Importantly, we find that across all three tasks, CreativityNeuro demonstrably reduces measures of mode collapse. Moreover, activation steering achieves comparable performance to CreativityNeuro on the DAT, but it does not transfer to the AUT and Task Task, demonstrating the effectiveness of weight-space steering in generalizing to unseen tasks. In conclusion, CreativityNeuro improves divergent thinking and reduces mode collapse without requiring behavioral data, re-training, or gradient-based fine-tuning, providing a straightforward way to enhance LLM performance in creative domains.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。