AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Learning Brain Representation with Hierarchical Visual Embeddings

Jiawen Zheng, Haonan Jia, MING LI, Yuhui Zheng, Yufeng Zeng, Yang Gao, Chen Liang

ICLR 2026 · Vision-language and multimodal › Human and embodied modeling · approach: fMRI visual decoding · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use hierarchical visual fusion with MBP over best prior methodUBP baseline · THINGS-EEG intra-subjecttop-1/top-5: 50.9/79.7 -> 75.7/94.6Improved
same, inter-subject EEGUBP baseline · THINGS-EEG inter-subjecttop-1/top-5: 12.4/33.4 -> 20.0/44.1Improved
same, MEG intra-subjectUBP baseline · THINGS-MEG intra-subjecttop-1/top-5: 26.7/55.2 -> 33.7/60.5Improved
same, MEG inter-subjectUBP baseline · THINGS-MEG inter-subjecttop-1/top-5: 2.2/10.4 -> 5.4/15.2Improved
add B32 to RN50 visual encoder (ablation)fusion retrieval · THINGS-EEG intra-subjecttop-1: 48.1 -> 56.9Improved
add VAE latent to B32 (ablation)fusion retrieval · THINGS-EEG intra-subjecttop-1: 52.2 -> 73.6Improved
add RN50 to B32+VAE (full triple) (ablation)fusion retrieval · THINGS-EEG intra-subjecttop-1: 73.6 -> 75.7Improved
add VAE to RN50 (ablation)fusion retrieval · THINGS-EEG inter-subjecttop-1: 12.7 -> 17.4Improved
add B32 to H14 fusion prior (ablation)reconstruction prior · EEG reconstructionCLIP: 0.773 -> 0.814Improved
add VAE to H14 fusion prior (ablation)reconstruction prior · EEG reconstructionCLIP: 0.773 -> 0.721Hurt
add VAE to H14+B32 fusion prior (ablation)reconstruction prior · EEG reconstructionPixCorr: 0.187 -> 0.195Mixed
use MBP instead of BrainProjection brain encoder (ablation)retrieval with full fusion · THINGS-EEG intra-subjecttop-1: 72.9 -> 75.7Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy