Research map / Vision-language and multimodal / Cross-modal learning
Shrinking the Teacher: An Adaptive Teaching Paradigm for Asymmetric EEG-Vision Alignment
Lukun Wu, Jie Li, Ziqi Ren, Kaifan Zhang, Xinbo Gao
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 |
|---|---|---|---|
| use ShrinkAdapter on CLIP features instead of no adapter (ablation) | ATS · THINGS-EEG zero-shot retrieval | avg top-1 accuracy (%): 51.20 -> 59.60 | Improved |
| use MLP or deep adapter instead of no adapter (ablation) | ATS · THINGS-EEG zero-shot retrieval | avg top-1 accuracy (%): 51.20 -> 47.40 (MLP), 46.05 (Deep) | Hurt |
| remove residual connection from ShrinkAdapter (ablation) | ShrinkAdapter (1:4) · THINGS-EEG | avg top-1 accuracy (%): 54.05 -> 59.60 | Improved |
| use 1:4 bottleneck compression instead of none (ablation) | ShrinkAdapter · THINGS-EEG | avg top-1 accuracy (%): 57.80 -> 59.60 | Improved |
| use 1:8 bottleneck instead of 1:4 (ablation) | ShrinkAdapter · THINGS-EEG | avg top-1 accuracy (%): 59.60 -> 56.05 | Hurt |
| increase latent dimension to 2048 (ablation) | ShrinkAdapter · THINGS-EEG | top-1 accuracy | No clear effect |
| add semantic distribution consistency loss with larger lambda (ablation) | ATS · THINGS-EEG | top-1 accuracy | Hurt |
| use STAE instead of EEGProject (ablation) | ATS brain encoder · THINGS-EEG | avg top-1 accuracy (%): 56.75 -> 60.20 | Improved |
| use stronger CLIP ViT-L/14 teacher instead of RN50 (ablation) | ATS · THINGS-EEG | top-1 accuracy: 10.0% lower than RN50 | Hurt |
| use ATS instead of UBP | THINGS-MEG retrieval · THINGS-MEG | top-1 accuracy: +3.0% | Improved |
Closest papers by meaning
- MindAlign: Bridging EEG, Vision, and Language for Zero-Shot Visual Decoding (NeurIPS 2026 · similarity 0.91)
- What Does the Brain See? Multiview Neural Representation to Demystify the Brain-Visual Alignment (NeurIPS 2026 · similarity 0.90)
- Neural‑Visual Decoding via Cognitive‑guided Adaptive Blurring and Information‑Constrained Alignment (NeurIPS 2026 · similarity 0.88)
- CognitionCapturer: Decoding Visual Stimuli from Human EEG Signal with Multimodal Information (AAAI 2025 · similarity 0.87)
- Correlating instruction-tuning (in multimodal models) with vision-language processing (in the brain) (ICLR 2025 · similarity 0.86)
- Human-Aligned Image Models Improve Visual Decoding from the Brain (ICML 2025 · similarity 0.86)
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.
