Research map / Vision-language and multimodal / Human and embodied modeling
CortiLife: A Unified Framework for Cortical Representation Learning across the Lifespan
Pengcheng Xue, Dong Nie, Jie Luo, Daoqiang Zhang, Xuyun Wen
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add the local topology encoder to the surface tokenizer (ablation) | CortiLife · CHD diagnosis (CT) | accuracy: 0.738 -> 0.806 | Improved |
| add the global interaction encoder to the surface tokenizer (ablation) | CortiLife · CHD diagnosis (CT) | accuracy: 0.740 -> 0.806 | Improved |
| add the patch-wise distribution (scale-adaptive) encoder (ablation) | CortiLife · CHD diagnosis (CT) | accuracy: 0.792 -> 0.806 | Improved |
| add the patch-wise distribution (scale-adaptive) encoder (ablation) | CortiLife · ADHD-200 diagnosis (CT) | accuracy: 0.615 -> 0.697 | Improved |
| add the scale-adaptive encoder before comparing age groups (ablation) | CortiLife · MMD ages 1-3 vs 6-66 years | maximum mean discrepancy: 0.852 -> 0.473 | Improved |
| use scale-adaptive encoder instead of LayerNorm+Linear on statistics (ablation) | CortiLife · ADHD-200 diagnosis (CT) | accuracy: 0.626 -> 0.697 | Improved |
| use scale-adaptive encoder instead of direct concatenation of mean and std (ablation) | CortiLife · ADHD-200 diagnosis (CT) | accuracy: 0.670 -> 0.697 | Improved |
| use scale-adaptive encoder instead of direct concatenation (MC modality) (ablation) | CortiLife · ADHD-200 diagnosis (MC) | AUC: 0.603 -> 0.618 | No clear effect |
| add metadata language prompting (L_CLIP loss) (ablation) | CortiLife · ADHD-200 diagnosis (CT) | AUC: 0.603 -> 0.730 | Improved |
| add age prompt to metadata prompting (ablation) | CortiLife · ADHD-200 diagnosis (CT) | accuracy: 0.660 -> 0.697 | Improved |
| add sex prompt to metadata prompting (ablation) | CortiLife · ADHD-200 diagnosis (CT) | accuracy: 0.687 -> 0.697 | Improved |
| use teacher-guided masking instead of random masking at ratio 0.75 (ablation) | CortiLife · ADHD-200 diagnosis (CT) | accuracy: 0.680 -> 0.697 | Improved |
Closest papers by meaning
- Brain Harmony: A Multimodal Foundation Model Unifying Morphology and Function into 1D Tokens (NeurIPS 2025 · similarity 0.85)
- Meta-Learning an In-Context Transformer Model of Human Higher Visual Cortex (NeurIPS 2025 · similarity 0.84)
- CLIP-MSM: A Multi-Semantic Mapping Brain Representation for Human High-Level Visual Cortex (AAAI 2025 · similarity 0.84)
- SIM: Surface-based fMRI Analysis for Inter-Subject Multimodal Decoding from Movie-Watching Experiments (ICLR 2025 · similarity 0.84)
- Finding Shared Decodable Concepts and their Negations in the Brain (ICLR 2025 · similarity 0.84)
- TopoNets: High performing vision and language models with brain-like topography (ICLR 2025 · similarity 0.83)
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.
