Come see our poster #61 at 3pm @iclr-conf.bsky.social #iclr2025 surface deep learning for generalising fMRI decoding across different human brains with @dahansimon.bsky.social @gabriben.bsky.social

Doing a demo on evaluating small language models for non-reasoning tasks (e.g. professionalize, shorten) at the amazon booth at 14:00 tomorrow at #ICLR 🇸🇬

This could allow to create personalised simulations of brain function for, say, visual impaired patients.
[3/3]

The popular aspirational goal in the field is to represent, in real time, what people are thinking as a movie. In this paper, we go beyond the fMRI -> image paradigm. With video+audio <-> fMRI we also predict which brain regions are activated across individuals.
[2/3]

SIM: Surface-based fMRI Analysis for Inter-Subject Multimodal Decoding from Movie-Watching Experiments Current AI frameworks for brain decoding and encoding, typically train and test models within the same datasets. This limits their utility for brain computer interfaces (BCI) or neurofeedback, for whi...

Excited to announce that our paper SIM: Surface-based fMRI Analysis for Inter-Subject Multimodal Decoding from Movie-Watching Experiments got accepted
@iclr_conf
🎉
arxiv.org/abs/2501.16471
[1/3]

[1/5] I'm very happy to share that our latest paper: “SIM: Surface-based fMRI Analysis for Inter-Subject Multimodal Decoding from Movie-Watching Experiments”, has been accepted to #ICLR2025! 🎉