Thomas Serre

Affiliations

Brown University

papers

874

total citations

papers (19)

CRAFT: Concept Recursive Activation FacTorization for Explainability

CVPR 2023arXiv

172

citations

What I Cannot Predict, I Do Not Understand: A Human-Centered Evaluation Framework for Explainability Methods

NEURIPS 2022arXiv

126

citations

Harmonizing the object recognition strategies of deep neural networks with humans

NEURIPS 2022arXiv

citations

A Holistic Approach to Unifying Automatic Concept Extraction and Concept Importance Estimation

NEURIPS 2023arXiv

citations

Look at the Variance! Efficient Black-box Explanations with Sobol-based Sensitivity Analysis

NEURIPS 2021arXiv

citations

Don't Lie to Me! Robust and Efficient Explainability With Verified Perturbation Analysis

CVPR 2023arXiv

citations

Performance-optimized deep neural networks are evolving into worse models of inferotemporal visual cortex

NEURIPS 2023arXiv

citations

Tracking Without Re-recognition in Humans and Machines

NEURIPS 2021arXiv

citations

Computing a human-like reaction time metric from stable recurrent vision models

NEURIPS 2023arXiv

citations

Diversity vs. Recognizability: Human-like generalization in one-shot generative models

NEURIPS 2022arXiv

citations

The 3D-PC: a benchmark for visual perspective taking in humans and machines

ICLR 2025arXiv

citations

Meta-Reinforcement Learning with Self-Modifying Networks

NEURIPS 2022arXiv

citations

Learning Functional Transduction

NEURIPS 2023arXiv

citations

Saliency strikes back: How filtering out high frequencies improves white-box explanations

ICML 2024arXiv

citations

Follow the Energy, Find the Path: Riemannian Metrics from Energy-Based Models

NEURIPS 2025arXiv

citations

Unlocking Feature Visualization for Deep Network with MAgnitude Constrained Optimization

NEURIPS 2023

citations

Thomas Serre

Affiliations

papers (19)

CRAFT: Concept Recursive Activation FacTorization for Explainability

What I Cannot Predict, I Do Not Understand: A Human-Centered Evaluation Framework for Explainability Methods

Harmonizing the object recognition strategies of deep neural networks with humans

A Holistic Approach to Unifying Automatic Concept Extraction and Concept Importance Estimation

Look at the Variance! Efficient Black-box Explanations with Sobol-based Sensitivity Analysis

A Benchmark for Compositional Visual Reasoning

Stable and expressive recurrent vision models

Break It Down: Evidence for Structural Compositionality in Neural Networks

Don't Lie to Me! Robust and Efficient Explainability With Verified Perturbation Analysis

Performance-optimized deep neural networks are evolving into worse models of inferotemporal visual cortex

Tracking Without Re-recognition in Humans and Machines

Computing a human-like reaction time metric from stable recurrent vision models

Diversity vs. Recognizability: Human-like generalization in one-shot generative models

The 3D-PC: a benchmark for visual perspective taking in humans and machines

Meta-Reinforcement Learning with Self-Modifying Networks

Learning Functional Transduction

Saliency strikes back: How filtering out high frequencies improves white-box explanations

Follow the Energy, Find the Path: Riemannian Metrics from Energy-Based Models

Unlocking Feature Visualization for Deep Network with MAgnitude Constrained Optimization

papers (19)

CRAFT: Concept Recursive Activation FacTorization for Explainability

What I Cannot Predict, I Do Not Understand: A Human-Centered Evaluation Framework for Explainability Methods

Harmonizing the object recognition strategies of deep neural networks with humans

A Holistic Approach to Unifying Automatic Concept Extraction and Concept Importance Estimation

Look at the Variance! Efficient Black-box Explanations with Sobol-based Sensitivity Analysis

A Benchmark for Compositional Visual Reasoning

Stable and expressive recurrent vision models

Break It Down: Evidence for Structural Compositionality in Neural Networks

Don't Lie to Me! Robust and Efficient Explainability With Verified Perturbation Analysis

Performance-optimized deep neural networks are evolving into worse models of inferotemporal visual cortex

Tracking Without Re-recognition in Humans and Machines

Computing a human-like reaction time metric from stable recurrent vision models

Diversity vs. Recognizability: Human-like generalization in one-shot generative models

The 3D-PC: a benchmark for visual perspective taking in humans and machines

Meta-Reinforcement Learning with Self-Modifying Networks

Learning Functional Transduction

Saliency strikes back: How filtering out high frequencies improves white-box explanations

Follow the Energy, Find the Path: Riemannian Metrics from Energy-Based Models

Unlocking Feature Visualization for Deep Network with MAgnitude Constrained Optimization