Research
The Language of Looking: Motion Tokens Reveal Domain-Dependent Local Structure in Gaze Sequences
arXiv:2610.16565v1 Announce Type: new Abstract: Human gaze is the primary high-bandwidth interface through which we parse the semantics of our environment. While often modeled as a sequence, the degre
arXiv:2610.16565v1 Announce Type: new Abstract: Human gaze is the primary high-bandwidth interface through which we parse the semantics of our environment. While often modeled as a sequence, the degree to which gaze exhibits transferable local structure across domains remains underexplored. Using motion tokenization as a diagnostic lens across four diverse datasets, we investigate the transferability of local sequence structure. We demonstrate that while gaze can share a reusable motion-token vocabulary, identical motion tokens are organized into different local transition patterns depending on the dataset and task context. Our results show the limitations of naive data-pooling under low-semantic motion-token representations and provide a framework for evaluating behavioral sequences through information-theoretic diagnostics.
Related
- Dense Local Dependencies Induce Attention-Logit Explosion and Training Instability During Long-Sequence Transformer Training
- Matrix-Decoupled Concentration for Autoregressive Sequences: Dimension-Free Guarantees for Sparse Long-Context Rewards
- Quantitative Clustering in Mean-Field Transformer Models
Source: arXiv cs.LG | 2026-10-14