Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Conformal Inference is (almost) Free for Neural Networks Trained with Early Stopping
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 CES instead of early stopping with data splitting | Neural networks for CIFAR10 outlier detection · CIFAR10 | TPR | Helped |
| Use CES instead of naive early stopping with the hybrid theoretical correction | Neural networks for CIFAR10 outlier detection · CIFAR10 | TPR | Helped |
| Use CES instead of full training without early stopping | Neural networks for CIFAR10 outlier detection · CIFAR10 | TPR | Helped |
| Use CES instead of other compared methods | Neural networks for CIFAR10 multi-class classification · CIFAR10 | Prediction-set cardinality | Helped |
| Use CES instead of early stopping with data splitting | Neural networks for regression · bio | Average interval width | Helped |
| Use full training instead of CES | Neural networks for regression · bio | Worst-slab conditional coverage | Hurt |
| Use full training instead of CES | Neural networks for quantile regression · homes | Worst-slab conditional coverage | Hurt |
| Use CES instead of early stopping with data splitting | Neural networks for quantile regression · homes | Worst-slab conditional coverage | Helped |
| Use the uncorrected naive benchmark instead of CES | Neural networks for outlier detection, classification, and regression · CIFAR10 and UCI datasets | Performance | No clear effect |
Closest papers by meaning
- Self-Calibrating Conformal Prediction (NeurIPS 2024 · similarity 0.89)
- Approximating Full Conformal Prediction for Neural Network Regression with Gauss-Newton Influence (ICLR 2025 · similarity 0.87)
- Leave-One-Out Stable Conformal Prediction (ICLR 2025 · similarity 0.87)
- Robust Conformal Outlier Detection under Contaminated Reference Data (ICML 2025 · similarity 0.87)
- Boosted Conformal Prediction Intervals (NeurIPS 2024 · similarity 0.87)
- Generalized Venn and Venn-Abers Calibration with Applications in Conformal Prediction (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.