Research map / Label-efficient and robust learning / Recognition and long-tail
When Deep Classifiers Agree: Analyzing Correlations between Learning Order and Image Statistics
Iuliia Pliushch, Martin Mundt, Nicolas Lupp, Visvanathan Ramesh
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 |
|---|---|---|---|
| track agreement over epochs against average image entropy of agreed instances | DenseNet121 · Pascal train | Pearson correlation: 0.88 | Improved |
| train with random labels instead of true labels (ablation) | DenseNet121 · CIFAR10 | agreement | Mixed |
| vary architecture (LeNet5, VGG16, ResNet50, DenseNet121) (ablation) | classifiers · CIFAR10 | agreement | No clear effect |
| use smaller batch sizes or simpler architectures (ablation) | classifiers · CIFAR10 | agreement speed | Mixed |
| compare object dataset Pascal to texture dataset KTH-TIPS2b for direction of correlation with entropy | DenseNet121 · Pascal vs KTH-TIPS2b | correlation direction | Mixed |
Closest papers by meaning
- Let’s Agree to Agree: Neural Networks Share Classification Order on Real Datasets (ICML 2020 · similarity 0.90)
- Quantifying Uncertainty in Error Consistency: Towards Reliable Behavioral Comparison of Classifiers (NeurIPS 2025 · similarity 0.85)
- Going Beyond Neural Network Feature Similarity: The Network Feature Complexity and Its Interpretation Using Category Theory (ICLR 2024 · similarity 0.85)
- A Decade's Battle on Dataset Bias: Are We There Yet? (ICLR 2025 · similarity 0.85)
- Objective drives the consistency of representational similarity across datasets (ICML 2025 · similarity 0.84)
- Directional Confusions Reveal Divergent Inductive Biases Through Rate-Distortion Geometry in Human and Machine Vision (NeurIPS 2026 · similarity 0.84)
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.
