大型语言模型在创造性思维过程中与人脑保持同步

Large Language Models Align with the Human Brain during Creative Thinking

April 3, 2026

作者: Mete Ismayilzada, Simone A. Luchini, Abdulkadir Gokce, Badr AlKhamissi, Antoine Bosselut, Antonio Laverghetta Jr., Lonneke van der Plas, Roger E. Beaty

cs.AI

摘要

創造性思維是人類認知的基本面向，而發散性思維——即產生新穎多元想法的能力——被普遍視為其核心生成引擎。大型語言模型近期在發散性思維測試中展現出卓越表現，先前研究亦表明，任務表現越優異的模型與人類大腦活動的對應程度越高。然而，現有的大腦-LLM對應研究多聚焦於被動性、非創造性的任務。本研究利用170名受試者執行「替代用途任務」時的fMRI數據，探討創造性思維過程中的大腦對應機制。我們從不同規模（2.7億至720億參數）的LLM提取表徵，並透過表徵相似性分析測量其與大腦反應的對應程度，重點關注與創造力相關的預設模式網絡和額頂網絡。研究發現：大腦-LLM對應程度隨模型規模（僅限預設模式網絡）與想法原創性（兩網絡皆然）提升，且此效應在創造過程初期最為顯著。我們進一步證實，訓練後目標會以功能選擇性方式塑造對應關係：經創造力優化的Llama-3.1-8B-Instruct模型能保持與高創造力神經反應的對應，同時降低與低創造力反應的對應；經人類行為微調的模型則能提升兩類對應；而推理訓練變體呈現相反模式，顯示思維鏈訓練會使表徵偏離創造性神經幾何結構，轉向分析性處理。這些結果表明，訓練後目標會選擇性地重塑LLM表徵，使其與人類創造性思維的神經幾何結構產生差異化對應。

English

Creative thinking is a fundamental aspect of human cognition, and divergent thinking-the capacity to generate novel and varied ideas-is widely regarded as its core generative engine. Large language models (LLMs) have recently demonstrated impressive performance on divergent thinking tests and prior work has shown that models with higher task performance tend to be more aligned to human brain activity. However, existing brain-LLM alignment studies have focused on passive, non-creative tasks. Here, we explore brain alignment during creative thinking using fMRI data from 170 participants performing the Alternate Uses Task (AUT). We extract representations from LLMs varying in size (270M-72B) and measure alignment to brain responses via Representational Similarity Analysis (RSA), targeting the creativity-related default mode and frontoparietal networks. We find that brain-LLM alignment scales with model size (default mode network only) and idea originality (both networks), with effects strongest early in the creative process. We further show that post-training objectives shape alignment in functionally selective ways: a creativity-optimized Llama-3.1-8B-Instruct preserves alignment with high-creativity neural responses while reducing alignment with low-creativity ones; a human behavior fine-tuned model elevates alignment with both; and a reasoning-trained variant shows the opposite pattern, suggesting chain-of-thought training steers representations away from creative neural geometry toward analytical processing. These results demonstrate that post-training objectives selectively reshape LLM representations relative to the neural geometry of human creative thought.