Robert Geirhos
18 papers · 2018–2026 · 5 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+9 more ↓ Show less ↑
π Interdisciplinary Bridge π Renaissance Researcher (5) π Conference Polyglot (5) π Academic Marathon (8) πΊοΈ Taxonomy Completionist (23)
π§
Keyword Pioneer
π£
Hot Topic Early Bird
π
Conference Polyglot
(5)
π₯
Mega-Team
(42)
β‘
Prolific Year
(5)
π
Century Club
(18)
π
Trend Setter
β
The Questioner
(3)
π₯
Unstoppable
(9)
Conferences
NIPS (6)
ICLR (5)
ICML (5)
CVPR (1)
WACV (1)
Top co-authors
Keywords
vision transformer
(3)
convolutional neural network
(3)
benchmark evaluation
(1)
object recognition
(1)
image retrieval
(1)
video generation
(1)
visual representation
(1)
foundation model
(1)
compositional learning
(1)
human behavior
(1)
visual object recognition
(1)
human visual system
(1)
model comparison
(1)
lifelong learning
(1)
self-supervised learning
(1)
contrastive learning
(1)
world model
(1)
feature visualization
(1)
model scaling
(1)
neural network architecture
(1)
Papers
Do Generative Video Models Understand Physical Principles?
WACV 2026
Position: We Canβt Understand AI Using our Existing Vocabulary
ICML 2025
LAION-C: An Out-of-Distribution Benchmark for Web-Scale Vision Models
ICML 2025
Towards flexible perception with visual memory
ICML 2025
Can We Talk Models Into Seeing the World Differently?
ICLR 2025
Learning Visual Composition through Improved Semantic Guidance
CVPR 2025
Intriguing Properties of Generative Classifiers
ICLR 2024
Donβt trust your eyes: on the (un)reliability of feature visualizations
ICML 2024
Scaling Vision Transformers to 22 Billion Parameters
ICML 2023
Patch nβ Pack: NaViT, a Vision Transformer for any Aspect Ratio and Resolution
NIPS 2023
Trivial or Impossible --- dichotomous data difficulty masks model differences (on ImageNet and beyond)
ICLR 2022
Beyond neural scaling laws: beating power law scaling via data pruning
NIPS 2022
Exemplary Natural Images Explain CNN Activations Better than State-of-the-Art Feature Visualization
ICLR 2021
Partial success in closing the gap between human and machine vision
NIPS 2021
How Well do Feature Visualizations Support Causal Understanding of CNN Activations?
NIPS 2021
Beyond accuracy: quantifying trial-by-trial behaviour of CNNs and humans by measuring error consistency
NIPS 2020
ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness
ICLR 2019
Generalisation in humans and deep neural networks
NIPS 2018