Quentin Bouniot

Postdoc@Helmholtz Zentrum München·德国·OpenReview

研究方向

uncertainty · explainability · calibration · object detection · self-supervised learning · semi-supervised learning · meta-learning · few-shot learning · optimal transport · deep learning · adversarial examples

4.5

TimeSAE: Sparse Decoding for Faithful Explanations of Black-Box Time Series Models

ICLR 2026Withdrawn

二作

4.0

Restyling Unsupervised Concept Based Interpretable Networks with Generative Models

ICLR 2025Poster

二作

3.0

Conceptualize Any Network: A Concept Extraction Framework for Holistic Interpretability of Image Classifiers

合作者 (20)

Quentin Bouniot

TimeSAE: Sparse Decoding for Faithful Explanations of Black-Box Time Series Models

Towards understanding multimodal in-context learning

TiViT: Time Series Representations Lie Hidden in Pretrained Vision Transformers

Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models

Tailoring Mixup to Data for Calibration

Restyling Unsupervised Concept Based Interpretable Networks with Generative Models

Conceptualize Any Network: A Concept Extraction Framework for Holistic Interpretability of Image Classifiers