What does Kiki look like? Cross-modal associations between speech sounds and visual shapes in vision-and-language models
This addresses the alignment of AI models with human cognitive biases, which is important for developing more human-like multimodal AI, though the findings are incremental as they primarily inform discussions rather than demonstrate a breakthrough.
The study investigated whether vision-and-language models (VLMs) exhibit the human-like bouba-kiki effect, a cross-modal association between speech sounds and visual shapes, by probing four VLMs but found no conclusive evidence, with results suggesting dependence on model features like architecture and training.
Humans have clear cross-modal preferences when matching certain novel words to visual shapes. Evidence suggests that these preferences play a prominent role in our linguistic processing, language learning, and the origins of signal-meaning mappings. With the rise of multimodal models in AI, such as vision- and-language (VLM) models, it becomes increasingly important to uncover the kinds of visio-linguistic associations these models encode and whether they align with human representations. Informed by experiments with humans, we probe and compare four VLMs for a well-known human cross-modal preference, the bouba-kiki effect. We do not find conclusive evidence for this effect but suggest that results may depend on features of the models, such as architecture design, model size, and training details. Our findings inform discussions on the origins of the bouba-kiki effect in human cognition and future developments of VLMs that align well with human cross-modal associations.