Research map / Label-efficient and robust learning / Recognition and long-tail
Rashomon Capacity: A Metric for Predictive Multiplicity in Classification
Hsiang Hsu, Flavio Calmon
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 AWP instead of random sampling to explore the Rashomon set | Rashomon Capacity estimation · UCI Adult, COMPAS, HSLS, CIFAR-10 | Rashomon Capacity | Improved |
| increase Rashomon parameter epsilon | Rashomon Capacity · all datasets | Rashomon Capacity | Improved |
| greedy selection of 10 models instead of all 163 models | Rashomon subset · COMPAS | Rashomon Capacity distribution: 163 -> 10 | No clear effect |
| ensemble methods | predictive multiplicity resolution · small models | Rashomon Capacity | Improved |
| weight regularization (LASSO or ridge) | logistic regression · logistic regression | Rashomon Capacity | Improved |
| perfect calibration of classifiers | classifiers · multiplicity | Rashomon Capacity | No clear effect |
| epsilon from 0.01 to 0.02 | Rashomon Capacity · UCI Adult vs COMPAS | Rashomon Capacity: 0.01 -> 0.02 | Mixed |
Closest papers by meaning
- Dropout-Based Rashomon Set Exploration for Efficient Predictive Multiplicity Estimation (ICLR 2024 · similarity 0.92)
- RashomonGB: Analyzing the Rashomon Effect and Mitigating Predictive Multiplicity in Gradient Boosting (NeurIPS 2024 · similarity 0.89)
- Position: Amazing Things Come From Having Many Good Models (ICML 2024 · similarity 0.87)
- ElliCE: Efficient and Provably Robust Algorithmic Recourse via the Rashomon Sets (NeurIPS 2025 · similarity 0.85)
- Accuracy vs. Accuracy: Computational Tradeoffs Between Classification Rates and Utility (NeurIPS 2026 · similarity 0.85)
- SORTeD Rashomon Sets of Sparse Decision Trees: Anytime Enumeration (NeurIPS 2025 · similarity 0.85)
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.
