mlsanityData to 14 Oct 2026

CP2: Leveraging Geometry for Conformal Prediction via Canonicalization

arXiv 2506.16189 ↗First posted 19 Jun 2025stat.ML cs.AI cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

We study the problem of conformal prediction (CP) under geometric data shifts, where data samples are susceptible to transformations such as rotations or flips. While CP endows prediction models with post-hoc uncertainty quantification and formal coverage guarantees, their practicality breaks under distribution shifts that deteriorate model performance. To address this issue, we propose integrating geometric information--such as geometric pose--into the conformal procedure to reinstate its guarantees and ensure robustness under geometric shifts. In particular, we explore recent advancements on pose canonicalization as a suitable information extractor for this purpose. Evaluating the combined approach across discrete and continuous shifts and against equivariant and augmentation-based baselines, we find that integrating geometric information with CP yields a principled way to address geometric shifts while maintaining broad applicability to black-box predictors.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 47 of its 64 references that are in the map. First posted on arXiv 19 Jun 2025.

Builds on

47 of 64 references.

References · 47

  1. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  2. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,223
  3. 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,375
  4. 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,287
  5. 2021Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gaugescited by 1,167
  6. 2016Group Equivariant Convolutional Networkscited by 1,165
  7. 2023Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMscited by 974
  8. 2021A Survey of Uncertainty in Deep Neural Networkscited by 680
  9. 2019Conformal Prediction Under Covariate Shiftcited by 519
  10. 2019General E(2)-Equivariant Steerable CNNscited by 446
  11. 2020Classification with Valid and Adaptive Coveragecited by 396
  12. 2021Adaptive Conformal Inference Under Distribution Shiftcited by 375
35 more
  1. 2022Conformal prediction beyond exchangeabilitycited by 355
  2. 2021Vector Neurons: A General Framework for SO(3)-Equivariant Networkscited by 305
  3. 2022Conformal Risk Controlcited by 299
  4. 2016Least Ambiguous Set-Valued Classifiers with Bounded Error Levelscited by 272
  5. 2019Gauge Equivariant Convolutional Networks and the Icosahedral CNNcited by 267
  6. 2020Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Datacited by 248
  7. 2021A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groupscited by 187
  8. 2021Frame Averaging for Invariant and Equivariant Network Designcited by 168
  9. 2022Adaptive Conformal Predictions for Time Seriescited by 161
  10. 2021Localized Conformal Prediction: A Generalized Inference Framework for Conformal Predictioncited by 114
  11. 2022Equivariance with Learned Canonicalization Functionscited by 109
  12. 2021Distribution-free uncertainty quantification for classification under label shiftcited by 101
  13. 2024Theoretical Foundations of Conformal Predictioncited by 101
  14. 2019B-Spline CNNs on Lie Groupscited by 97
  15. 2023Conformal PID Control for Time Series Predictioncited by 88
  16. 2021Conformal Prediction using Conditional Histogramscited by 77
  17. 2023FAENet: Frame Averaging Equivariant GNN for Materials Modelingcited by 73
  18. 2022Batch Multivalid Conformal Predictioncited by 67
  19. 2024Online conformal prediction with decaying step sizescited by 56
  20. 2023Clifford Group Equivariant Neural Networkscited by 55
  21. 2024Does equivariance matter at scale?cited by 52
  22. 2019With Malice Towards None: Assessing Uncertainty via Equalized Coveragecited by 43
  23. 2024Equivariant Frames and the Impossibility of Continuous Canonicalizationcited by 43
  24. 2023Learning Probabilistic Symmetrization for Architecture Agnostic Equivariancecited by 26
  25. 2024Conformal Validity Guarantees Exist for Any Data Distribution (and How to Find Them)cited by 26
  26. 2024Adaptive Bounding Box Uncertainties via Two-Step Conformal Predictioncited by 23
  27. 2022Equi-Tuning: Group Equivariant Fine-Tuning of Pretrained Modelscited by 22
  28. 2020Empirical Frequentist Coverage of Deep Learning Uncertainty Quantification Procedurescited by 19
  29. 2023Adaptive Conformal Prediction by Reweighting Nonconformity Scorecited by 16
  30. 2023Efficient Equivariant Transfer Learning from Pretrained Modelscited by 11
  31. 2024A Generative Model of Symmetry Transformationscited by 10
  32. 2024Improved Canonicalization for Model Agnostic Equivariancecited by 7
  33. 2024Learning Symmetries via Weight-Sharing with Doubly Stochastic Tensorscited by 5
  34. 2023Self-Supervised Detection of Perfect and Partial Input-Dependent Symmetriescited by 4
  35. 2024Conformal time series decomposition with component-wise exchangeabilitycited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025On Equivariant Model Selection through the Lens of Uncertaintycited by 1
  2. 2026eCP: Equivariant Conformal Prediction with pre-trained modelscited by 0