mlsanityData to 7 Oct 2026

Prospective Learning: Principled Extrapolation to the Future

arXiv 2201.07372 ↗First posted 19 Jan 2022cs.LG cs.AIcited by 6

Not on the map. Why?

Abstract (arXiv)

Learning is a process which can update decision rules, based on past experience, such that future performance improves. Traditionally, machine learning is often evaluated under the assumption that the future will be identical to the past in distribution or change adversarially. But these assumptions can be either too optimistic or pessimistic for many problems in the real world. Real world scenarios evolve over multiple spatiotemporal scales with partially predictable dynamics. Here we reformulate the learning problem to one that centers around this idea of dynamic futures that are partially learnable. We conjecture that certain sequences of tasks are not retrospectively learnable (in which the data distribution is fixed), but are prospectively learnable (in which distributions may be dynamic), suggesting that prospective learning is more difficult in kind than retrospective learning. We argue that prospective learning more accurately characterizes many real world problems that (1) currently stymie existing artificial intelligence solutions and/or (2) lack adequate explanations for how natural intelligences solve them. Thus, studying prospective learning will lead to deeper insights and solutions to currently vexing challenges in both natural and artificial intelligences.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 20 of its 57 references that are in the map. First posted on arXiv 19 Jan 2022.

Builds on

20 of 57 references.

References · 20

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  3. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  4. 2019Three scenarios for continual learningcited by 610
  5. 2019Generalizing from a Few Examples: A Survey on Few-Shot Learningcited by 462
  6. 2021Towards Out-Of-Distribution Generalization: A Surveycited by 431
  7. 2019Online Learning: A Modern Introduction Using Convex Optimizationcited by 407
  8. 2019Meta-Dataset: A Dataset of Datasets for Learning to Learn from Few Examplescited by 312
  9. 2018Meta-Learning: A Surveycited by 254
  10. 2021Mind the Gap: Assessing Temporal Generalization in Neural Language Modelscited by 226
  11. 2018Online Learning: A Comprehensive Surveycited by 203
  12. 2019REMIND Your Neural Network to Prevent Catastrophic Forgettingcited by 197
8 more
  1. 2018Strike (with) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objectscited by 143
  2. 2018A Tour of Reinforcement Learning: The View from Continuous Controlcited by 129
  3. 2021OoD-Bench: Quantifying and Understanding Two Dimensions of Out-of-Distribution Generalizationcited by 86
  4. 2021Out of Distribution Generalization in Machine Learningcited by 59
  5. 2021Model Zoo: A Growing "Brain" That Learns Continuallycited by 53
  6. 2020Meta-learning in natural and artificial intelligencecited by 39
  7. 2020Online Structured Meta-learningcited by 27
  8. 2021Towards a theory of out-of-distribution learningcited by 11

Built on by

Cited by year

Citing papers per year: 2024: 2; 2025: 4.

6 papers.

  1. 2025Latent learning: episodic memory complements parametric learning by enabling flexible reuse of experiencescited by 14
  2. 2024Transformers for Supervised Online Continual Learningcited by 5
  3. 2024Prospective Learning: Learning for a Dynamic Futurecited by 4
  4. 2025Causality in the human niche: lessons for machine learningcited by 1
  5. 2025Prospective Learning in Retrospectcited by 1
  6. 2025Learning Model Successorscited by 0