Research map / Adversarial robustness and security / Attacks on text and graphs
Alignment with human representations supports robust few-shot learning
Ilia Sucholutsky, Thomas L. Griffiths
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 |
|---|---|---|---|
| select the 5 most Pearson-aligned models instead of the 5 closest to the mean | pretrained vision models · ImageNet-A top-1 | accuracy: 10.26 -> 59.93 | Improved |
| select the 5 least-aligned models instead of the 5 mid-aligned models | pretrained vision models · ImageNet-A top-1 | accuracy: 10.26 -> 30.89 | Improved |
| select the 5 most Pearson-aligned models instead of the 5 closest to the mean | pretrained vision models · ImageNet-R top-1 | accuracy: 39.16 -> 61.57 | Improved |
| select the 5 most Pearson-aligned models instead of the 5 closest to the mean | pretrained vision models · ImageNet-S top-1 | accuracy: 26.90 -> 48.06 | Improved |
| select the 5 most triplet-aligned models instead of the 5 closest to the mean | pretrained vision models · ImageNet-A top-5 | accuracy: 46.63 -> 72.20 | Improved |
| use z^2 alignment instead of raw alignment to predict robustness | alignment metrics · ImageNet-A | correlation | Mixed |
| measure alignment on dissimilar data (MNIST instead of CIFAR100) | Pearson z^2 alignment · MNIST few-shot | correlation with accuracy | Hurt |
| add ImageNet-21k/22k pre-training before ImageNet-1k fine-tuning | vision models · alignment and few-shot | alignment and accuracy | Improved |
| apply knowledge distillation | vision models · alignment and few-shot | alignment and accuracy | Improved |
| add batch normalization to VGG | VGG · alignment and few-shot | alignment | Mixed |
| prune EfficientNet | EfficientNet · alignment and few-shot | alignment | Mixed |
| increase ResNet depth | ResNet · alignment and few-shot | alignment | Mixed |
Closest papers by meaning
- Not Too Generative, Not Too Discriminative: The Human Alignment Sweet Spot (NeurIPS 2026 · similarity 0.91)
- Towards a learning theory of representation alignment (ICLR 2025 · similarity 0.88)
- Evaluating alignment between humans and neural network representations in image-based learning tasks (NeurIPS 2024 · similarity 0.88)
- Learning Human-like Representations to Enable Learning Human Values (NeurIPS 2024 · similarity 0.87)
- When does perceptual alignment benefit vision representations? (NeurIPS 2024 · similarity 0.86)
- Position: Towards Bidirectional Human-AI Alignment (NeurIPS 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.
