Interoceptive Divergence in Aesthetic Evaluation and Implications for Human-AI Alignment

• URL: http://arxiv.org/abs/2605.18759v1
• Date: Sun, 05 Apr 2026 10:20:46 GMT
• Title: Interoceptive Divergence in Aesthetic Evaluation and Implications for Human-AI Alignment
• Abstract summary: This study investigates how humans and AI systems converge or diverge in aesthetic experiences.<n>While humans and AI exhibited broadly similar patterns in the correlations between beauty ratings and emotions, notable divergences emerged in both the distribution of emotional responses and the relationship between beauty ratings and bodily sensations.<n>Findings suggest that state-of-the-art LLMs, trained on large-scale textual data, can approximate average human tendencies in aesthetic evaluation to a certain extent.
• Score: 2.905751301655124
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Artificial intelligence (AI), exemplified by large language models (LLMs), is rapidly approaching and in some cases surpassing human performance across a wide range of cognitive tasks. However, human nature is not limited to intelligence alone; it also encompasses sensibility, including the capacity to perceive and experience beauty in visual scenes. This raises a fundamental question: how humans and AI systems converge or diverge in such aesthetic experiences. Aesthetic evaluation depends not only on objective properties of images but also on internal processes within the observer. As part of ongoing efforts in AI alignment, building upon prior human studies that have examined the relationship between beauty ratings, bodily sensations, and emotions, we adopt a comparable set of questionnaire items and present them to LLMs, enabling a direct comparison between human and AI responses. Our comparative analyses revealed that, while humans and AI exhibited broadly similar patterns in the correlations between beauty ratings and emotions, as well as in the image features they prioritized, notable divergences emerged in both the distribution of emotional responses and the relationship between beauty ratings and bodily sensations. These findings suggest that state-of-the-art LLMs, trained on large-scale textual data, can approximate average human tendencies in aesthetic evaluation to a certain extent. However, they also indicate limitations, particularly in relation to interoceptive aspects, which may reflect insufficient representation in training data or unintended consequences of alignment processes. These findings highlight key challenges for AI alignment and suggest important directions for developing AI systems with human-like aesthetic processing.

Related papers

• Imperfectly Cooperative Human-AI Interactions: Comparing the Impacts of Human and AI Attributes in Simulated and User Studies [25.913918889459552]
  AI design characteristics and human personality traits each impact the quality and outcomes of human-AI interactions.<n>This study compares a purely simulated dataset comprising 2,000 simulations and a parallel human subjects experiment involving 290 human participants.<n>Results reveal divergences between purely simulated and human study datasets, and between scenario types.
  arXiv  Detail & Related papers  (2026-04-17T01:10:34Z)
• HumanPCR: Probing MLLM Capabilities in Diverse Human-Centric Scenes [72.26829188852139]
  HumanPCR is an evaluation suite for probing MLLMs' capacity about human-related visual contexts.<n>Human-P, HumanThought-C, and Human-R feature over 6,000 human-verified multiple choice questions.<n>Human-R offers a challenging manually curated video reasoning test.
  arXiv  Detail & Related papers  (2025-08-19T09:52:04Z)
• Artificial Intelligence Can Emulate Human Normative Judgments on Emotional Visual Scenes [0.09208007322096533]
  We study whether state-of-the-art multimodal systems can emulate human emotional ratings on a standardized set of images.<n>The AI judgements correlate surprisingly well with the average human ratings.
  arXiv  Detail & Related papers  (2025-03-24T15:41:23Z)
• Identifying Features that Shape Perceived Consciousness in Large Language Model-based AI: A Quantitative Study of Human Responses [4.369058206183195]
  This study quantitively examines which features of AI-generated text lead humans to perceive subjective consciousness in large language model (LLM)-based AI systems.<n>Using regression and clustering analyses, we investigated how these features influence participants' perceptions of AI consciousness.
  arXiv  Detail & Related papers  (2025-02-21T10:27:28Z)
• AI-generated Image Quality Assessment in Visual Communication [72.11144790293086]
  AIGI-VC is a quality assessment database for AI-generated images in visual communication.<n>The dataset consists of 2,500 images spanning 14 advertisement topics and 8 emotion types.<n>It provides coarse-grained human preference annotations and fine-grained preference descriptions, benchmarking the abilities of IQA methods in preference prediction, interpretation, and reasoning.
  arXiv  Detail & Related papers  (2024-12-20T08:47:07Z)
• MEMO-Bench: A Multiple Benchmark for Text-to-Image and Multimodal Large Language Models on Human Emotion Analysis [53.012111671763776]
  This study introduces MEMO-Bench, a comprehensive benchmark consisting of 7,145 portraits, each depicting one of six different emotions. Results demonstrate that existing T2I models are more effective at generating positive emotions than negative ones. Although MLLMs show a certain degree of effectiveness in distinguishing and recognizing human emotions, they fall short of human-level accuracy.
  arXiv  Detail & Related papers  (2024-11-18T02:09:48Z)
• How explainable AI affects human performance: A systematic review of the behavioural consequences of saliency maps [0.0]
  Saliency maps can explain how deep neural networks classify images. But are they actually useful for humans? The present systematic review of 68 user studies found that while saliency maps can enhance human performance, null effects or even costs are quite common.
  arXiv  Detail & Related papers  (2024-04-03T21:46:25Z)
• Exploring the Naturalness of AI-Generated Images [59.04528584651131]
  We take the first step to benchmark and assess the visual naturalness of AI-generated images. We propose the Joint Objective Image Naturalness evaluaTor (JOINT), to automatically predict the naturalness of AGIs that aligns human ratings. We demonstrate that JOINT significantly outperforms baselines for providing more subjectively consistent results on naturalness assessment.
  arXiv  Detail & Related papers  (2023-12-09T06:08:09Z)
• Data-driven emotional body language generation for social robotics [58.88028813371423]
  In social robotics, endowing humanoid robots with the ability to generate bodily expressions of affect can improve human-robot interaction and collaboration. We implement a deep learning data-driven framework that learns from a few hand-designed robotic bodily expressions. The evaluation study found that the anthropomorphism and animacy of the generated expressions are not perceived differently from the hand-designed ones.
  arXiv  Detail & Related papers  (2022-05-02T09:21:39Z)
• Affect-DML: Context-Aware One-Shot Recognition of Human Affect using Deep Metric Learning [29.262204241732565]
  Existing methods assume that all emotions-of-interest are given a priori as annotated training examples. We conceptualize one-shot recognition of emotions in context -- a new problem aimed at recognizing human affect states in finer particle level from a single support sample. All variants of our model clearly outperform the random baseline, while leveraging the semantic scene context consistently improves the learnt representations.
  arXiv  Detail & Related papers  (2021-11-30T10:35:20Z)
• Affective Image Content Analysis: Two Decades Review and New Perspectives [132.889649256384]
  We will comprehensively review the development of affective image content analysis (AICA) in the recent two decades. We will focus on the state-of-the-art methods with respect to three main challenges -- the affective gap, perception subjectivity, and label noise and absence. We discuss some challenges and promising research directions in the future, such as image content and context understanding, group emotion clustering, and viewer-image interaction.
  arXiv  Detail & Related papers  (2021-06-30T15:20:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.