Research map / Label-efficient and robust learning / Recognition and long-tail
Local Intrinsic Dimension of Representations Predicts Alignment and Generalization in AI Models and Human Brain
Junjie Yu, Wenxiao Ma, Chen Wei, Jianyu Zhang, Haotian Deng, Zihan Deng, Quanying Liu
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 lower local intrinsic dimension (K=100) | vision models · ConvNeXt models, NSD + ImageNet-1K | AI-AI alignment | Improved |
| lower local intrinsic dimension | vision models · NSD EBA region | AI-Brain alignment | Improved |
| lower local intrinsic dimension | vision models · ImageNet-1K | generalization | Improved |
| measure dimension at global scale (large K) instead of local (ablation) | intrinsic dimension · ConvNeXt models | correlation with alignment and generalization | Hurt |
| randomly subsample 1,000 embeddings globally instead of local neighborhood (ablation) | intrinsic dimension · ConvNeXt models | correlation with alignment | Hurt |
| increase model size | ConvNeXt · NSD, ImageNet-1K | local intrinsic dimension | Improved |
| increase pretraining dataset size | ConvNeXt · NSD, ImageNet-1K | local intrinsic dimension | Improved |
| pool different architectures (ablation) | intrinsic dimension correlations · ViT, ConvNeXt, ResNet, ResMLP | correlation strength | Mixed |
| architecture family as a factor | AI-AI alignment · cross-architecture | inter-model alignment | Mixed |
Closest papers by meaning
- Bridging Critical Gaps in Convergent Learning: How Representational Alignment Evolves Across Layers, Training, and Distribution Shifts (NeurIPS 2025 · similarity 0.90)
- Evaluating alignment between humans and neural network representations in image-based learning tasks (NeurIPS 2024 · similarity 0.89)
- Towards a learning theory of representation alignment (ICLR 2025 · similarity 0.88)
- Model–Behavior Alignment under Flexible Evaluation: When the Best-Fitting Model Isn’t the Right One (NeurIPS 2025 · similarity 0.88)
- Scaling and context steer LLMs along the same computational path as the human brain (NeurIPS 2025 · similarity 0.87)
- Scaling Laws for Task-Optimized Models of the Primate Visual Ventral Stream (ICML 2025 · similarity 0.87)
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.
