Research map / Point clouds / Point cloud classification
Representer Point Selection for Explaining Deep Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 representer values instead of influence values to rank mislabeled training examples for inspection | CIFAR-10 binary automobile-versus-horse logistic regression · CIFAR-10 | test accuracy after dataset correction | Helped |
| Use representer values instead of random selection to rank mislabeled training examples | CIFAR-10 binary automobile-versus-horse logistic regression · CIFAR-10 | test accuracy after dataset correction | Helped |
| Use representer points rather than influence functions to produce positive and negative examples | Pretrained ResNet-50 fine-tuned on AwA · AwA | usefulness of retrieved examples | Helped |
| Use representer values rather than influence values to compute training-point explanations | CIFAR-10 and AwA explanation runs · CIFAR-10 and AwA | computation time: 267.08 \pm 248.20 -> 0.10 \pm 0.08; 172.71 \pm 32.63 -> 0.19 \pm 0.12 | Helped |
| Use representer values rather than influence values to select influential training points | CIFAR-10 test images · CIFAR-10 | availability of non-trivial training-point explanations | Helped |
| Use L2-regularized representer values instead of influence functions | 2-D toy binary-classification data · 2-D toy data | availability of positive and negative examples | Helped |
Closest papers by meaning
- Explaining, Fast and Slow: Abstraction and Refinement of Provable Explanations (ICML 2025 · similarity 0.84)
- Parameter symmetries determine representational geometry in overparameterized nonlinear networks (NeurIPS 2026 · similarity 0.84)
- When Representations Align: Universality in Representation Learning Dynamics (ICML 2024 · similarity 0.83)
- Explainable Neural Networks with Guarantee: A Sparse Estimation Approach (AAAI 2025 · similarity 0.83)
- Formation of Representations in Neural Networks (ICLR 2025 · similarity 0.83)
- Expand-and-Cluster: Parameter Recovery of Neural Networks (ICML 2024 · 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.