Source of the temporal-context benefit

Establish whether NOEMA’s benefit from temporal context arises from accommodation of the delayed hemodynamic response, stimulus-related temporal structure, or both.

Background

The experiments show that NOEMA benefits from multiple consecutive fMRI frames and from preserving their chronological order, outperforming a Context MLP with the same temporal input. However, the observed advantage could arise from different mechanisms.

Specifically, temporal modeling may help compensate for fMRI’s delayed hemodynamic response, exploit temporal structure in the speech stimulus, or combine both effects. The reported ablations do not distinguish these explanations.

References

Our analyses also do not establish what underlies the benefit of temporal context: it may reflect accommodation of the delayed hemodynamic response, stimulus-related temporal structure, or both.

— Decoding Naturalistic Speech from fMRI with Temporal Modeling and Multimodal Alignment  (2610.22049 - Caris et al., 15 Oct 2026) in Section Conclusion, subsection “Limitations and future work”