What Do Language Models Hear? Probing for Auditory Representations in Language Models
This addresses the problem of understanding grounded knowledge in language models for researchers in AI and cognitive science, but it is incremental as it builds on existing probing techniques.
The study investigated whether language models encode meaningful auditory representations of objects, finding that a linear probe trained with contrastive loss could generalize to unseen objects with above-chance accuracy across various models.
This work explores whether language models encode meaningfully grounded representations of sounds of objects. We learn a linear probe that retrieves the correct text representation of an object given a snippet of audio related to that object, where the sound representation is given by a pretrained audio model. This probe is trained via a contrastive loss that pushes the language representations and sound representations of an object to be close to one another. After training, the probe is tested on its ability to generalize to objects that were not seen during training. Across different language models and audio models, we find that the probe generalization is above chance in many cases, indicating that despite being trained only on raw text, language models encode grounded knowledge of sounds for some objects.