Code for the Conditional Mutual Information-Debiasing (CMID) method.

[WACV 2024] Source code for "Consolidating separate degradations model via weights fusion and distillation".

[WACV 2024] Domain Generalisation via Risk Distribution Matching

Official code for the paper: "When and How Does CLIP Enable Domain and Compositional Generalization?" (ICML 2025 Spotlight)

[ICML 2025] OOD-Chameleon: Is Algorithm Selection for OOD Generalization Learnable?

Möbius Transform for Mitigating Perspective Distortions in Representation Learning (ECCV 2024)

Official code for "Tropical Attention: Neural Algorithmic Reasoning for Combinatorial Algorithms"

Source Code for Our ICML-2025 Paper "Controlling Neural Collapse Enhances Out-of-Distribution Detection and Transfer Learning"

Evaluation of non-ROI masking to improve OOD generalization in chest x-ray disease classification

The value of out-of-distribution data (ICML 2023)

Can Self-Supervised Representation Learning Methods Withstand Distribution Shifts and Corruptions? (ICCV Workshops 2023)

Mitigating Spurious Correlations for Self-supervised Recommendation

[CVPR 2024] Source code for "Diffusion-Based Adaptation for Classification of Unknown Degraded Images".

Source Code for Our NeurIPS-2024 Paper "What Variables Affect Out-of-Distribution Generalization in Pretrained Models?"

[ECCV Workshops 2024] BelHouse3D: A Benchmark Dataset for Assessing Occlusion Robustness in 3D Point Cloud Semantic Segmentation

Code for the ICML 2021 paper "Composed Fine-Tuning: Freezing Pre-Trained Denoising Autoencoders for Improved Generalization" by Sang Michael Xie, Tengyu Ma, Percy Liang

Code for "Environment Diversification with Multi-head Neural Network for Invariant Learning" (NeurIPS 2022)

Masking Strategies for Background Bias Removal in Computer Vision Models (ICCVW OODCV 2023 paper)

Implementation of the paper SAM-Deblur: Let Segment Anything Boost Image Deblurring(ICASSP2024)

This repository contains the ViewFool and ImageNet-V proposed by the paper “ViewFool: Evaluating the Robustness of Visual Recognition to Adversarial Viewpoints” (NeurIPS2022).