GPT-4.1 能复现心理化的语言结构：LLM“算法反思性”研究
Algorithmic Reflectivity in Large Language Models: Methodological Study of Linguistic Structure in Mentalization
研究用 GPT-4.1 在标准化条件下生成 50 段人机对话及叙事性心理化画像，5 名受训精神科医生按 4 个 MBT 轴和 2 个高阶维度独立评分，均分 3.63-3.98，评估域 ICC(3,k) 为 0.60-0.84。
Authors:Stefano Epifani (1 and 2), Giuliano Castigliego (1 and 3), Giuliano Razzicchia (1), Elisabeth Seiwald-Sonderegger (4), Laura Kecskeméti (4) ((1) Digital Transformation Institute, Rome, Italy, (2) Department of Communication and Social Research, Sapienza University of Rome, Italy, (3) Psychoanalytic Academy of Italian-Speaking Switzerland, Chur, Switzerland, (4) Psychiatric Services of the Canton of Grisons, Chur, Switzerland)
Abstract:Mentalization is a core human capacity involving the interpretation of one's own and others' behavior in terms of underlying mental states. Large language models (LLMs) can generate linguistically reflective discourse, raising the question of whether the formal linguistic structure of mentalization can be reproduced without experiential and affective processes. This study assessed whether an LLM can generate mentalization profiles structurally coherent with the dimensions of mentalization-based treatment (MBT) and recognized as formally mentalizing by expert clinicians. Fifty human-model dialogues were generated with GPT-4.1 under standardized conditions; at the end of each, the model produced a narrative mentalization profile. Five MBT-trained psychiatrists independently rated each profile on 4 canonical MBT axes and 2 higher-order dimensions (regulatory functions, synthetic profile) for evaluative and argumentative coherence. Interrater reliability was assessed with ICC(3,k). Mean scores ranged from 3.63 to 3.98, indicating moderate-to-high structural coherence; ICC(3,k) in the evaluative domain ranged from 0.60 to 0.84. Evaluative and argumentative coherence were strongly associated (Pearson r 0.826-0.967). A negative control with mismatched dialogue-profile pairs yielded minimal ratings. Clinicians described outputs as linguistically organized and clinically interpretable, but with limited affective variability and contextual integration. The findings suggest that LLMs can reproduce the formal linguistic structure of reflective discourse - a phenomenon we term algorithmic reflectivity - which appears more consistent with structural language regularities than with experiential or intentional processes. LLMs may serve as research and training tools for studying the linguistic aspects of mentalization.
| Comments: | Published version in JMIR AI 2026;5:e90750, this https URL. Licensed under Creative Commons Attribution cc-by 4.0. 3 tables, 1 figure. Previous title: "The Linguistic Architecture of Reflective Thought: Evaluation of a Large Language Model as a Tool to Isolate the Formal Structure of Mentalization" |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2512.08945 [cs.CL] |
| (or arXiv:2512.08945v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2512.08945 arXiv-issued DOI via DataCite |
|
| Journal reference: | JMIR AI 2026;5:e90750 |
| Related DOI: | https://doi.org/10.2196/90750
DOI(s) linking to related resources |
Submission history
From: Stefano Epifani [view email]
[v1]
Thu, 20 Nov 2025 23:51:34 UTC (260 KB)
[v2]
Tue, 20 Oct 2026 21:10:31 UTC (313 KB)
来源：arXiv cs.CL · arxiv.org