mlsanityData to 7 Oct 2026

Alignment with human representations supports robust few-shot learning

arXiv 2301.11990 ↗First posted 27 Jan 2023cs.LG cs.AI cs.CVcited by 25

Not on the map. Why?

Abstract (arXiv)

Should we care whether AI systems have representations of the world that are similar to those of humans? We provide an information-theoretic analysis that suggests that there should be a U-shaped relationship between the degree of representational alignment with humans and performance on few-shot learning tasks. We confirm this prediction empirically, finding such a relationship in an analysis of the performance of 491 computer vision models. We also show that highly-aligned models are more robust to both natural adversarial attacks and domain shifts. Our results suggest that human-alignment is often a sufficient, but not necessary, condition for models to make effective use of limited data, be robust, and generalize well.

Key facts. Cited by 25 papers in the map (data through 7 Oct 2026). Builds on 27 of its 43 references that are in the map. First posted on arXiv 27 Jan 2023.

Builds on

27 of 43 references.

References · 27

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  4. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  5. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  6. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  7. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  8. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  9. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
  10. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
  11. 2021BEiT: BERT Pre-Training of Image Transformerscited by 2,360
  12. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
15 more
  1. 2019Natural Adversarial Examplescited by 1,401
  2. 2021Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gaugescited by 1,131
  3. 2019Learning Robust Global Representations by Penalizing Local Predictive Powercited by 1,010
  4. 2022Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distributioncited by 722
  5. 2020Artificial Intelligence, Values and Alignmentcited by 443
  6. 2021High-Performance Large-Scale Image Recognition Without Normalizationcited by 359
  7. 2022Relative representations enable zero-shot latent space communicationcited by 175
  8. 2021Revisiting ResNets: Improved Training and Scaling Strategiescited by 158
  9. 2022Human alignment of neural network representationscited by 74
  10. 2022Harmonizing the object recognition strategies of deep neural networks with humanscited by 59
  11. 2022Do better ImageNet classifiers assess perceptual similarity better?cited by 34
  12. 2019Learning as the Unsupervised Alignment of Conceptual Systemscited by 14
  13. 2022On the Informativeness of Supervision Signalscited by 12
  14. 2022Predicting Human Similarity Judgments Using Large Language Modelscited by 10
  15. 2022Researching Alignment Research: Unsupervised Analysiscited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 7; 2024: 11; 2025: 4; 2026: 3.

25 papers.

Sort:

By area:

  1. 2023Getting aligned on representational alignmentcited by 142
  2. 2024Building Machines that Learn and Think with Peoplecited by 73
  3. 2023Improving neural network representations using human similarity judgmentscited by 38
  4. 2023Performance-optimized deep neural networks are evolving into worse models of inferotemporal visual cortexcited by 18
  5. 2024Position: Towards Bidirectional Human-AI Alignmentcited by 15
  6. 2023What Matters to You? Towards Visual Representation Alignment for Robot Learningcited by 14
  7. 2023Adversarial alignment: Breaking the trade-off between the strength of an attack and its relevance to human perceptioncited by 7
  8. 2023Learning Human-like Representations to Enable Learning Human Valuescited by 7
  9. 2024How Aligned are Different Alignment Metrics?cited by 6
  10. 2024Maximizing Alignment with Minimal Feedback: Efficiently Learning Rewards for Visuomotor Robot Policy Alignmentcited by 6
  11. 2024Improving deep learning with prior knowledge and cognitive models: A survey on enhancing explainability, adversarial robustness and zero-shot learningcited by 4
  12. 2024TurtleBench: A Visual Programming Benchmark in Turtle Geometrycited by 4
13 more
  1. 2023Dimensions of Disagreement: Unpacking Divergence and Misalignment in Cognitive Science and Artificial Intelligencecited by 3
  2. 2024Representational Alignment Supports Effective Machine Teachingcited by 3
  3. 2024Knowledge graphs for empirical concept retrievalcited by 2
  4. 2024A Nurse is Blue and Elephant is Rugby: Cross Domain Alignment in Large Language Models Reveal Human-like Patternscited by 2
  5. 2025Alignment and Adversarial Robustness: Are More Human-Like Models More Secure?cited by 2
  6. 2024Connecting Concept Convexity and Human-Machine Alignment in Deep Neural Networkscited by 1
  7. 2025Geometry matters: insights from Ollivier Ricci Curvature and Ricci Flow into representational alignment through Ollivier-Ricci Curvature and Ricci Flowcited by 1
  8. 2025Uncovering the Computational Ingredients of Human-Like Representations in LLMscited by 1
  9. 2026Learning to Communicate Across Modalities: Perceptual Heterogeneity in Multi-Agent Systemscited by 1
  10. 2024Building an Ethical and Trustworthy Biomedical AI Ecosystem for the Translational and Clinical Integration of Foundational Modelscited by 0
  11. 2025Replicating Human Social Perception in Generative AI: Evaluating the Valence-Dominance Modelcited by 0
  12. 2026Gaslight, Gatekeep, V1-V3: Early Visual Cortex Alignment Shields Vision-Language Models from Sycophantic Manipulationcited by 0
  13. 2026Capability neq Interpretability: Human Interpretability of Vision Foundation Modelscited by 0