paper-with-me

홈 › Papers

Adversarial Support Alignment

2022-03-16 · ICLR 2022 4 · Shangyuan Tong, Timur Garipov, Yang Zhang, Shiyu Chang, Tommi S. Jaakkola

We study the problem of aligning the supports of distributions. Compared to the existing work on distribution alignment, support alignment does not require the densities to be matched. We propose symmetric support difference as a divergence measure to quantify the mismatch between supports. We show that select discriminators (e.g. discriminator trained for Jensen-Shannon divergence) are able to map support differences as support differences in their one-dimensional output space. Following this result, our method aligns supports by minimizing a symmetrized relaxed optimal transport cost in the discriminator 1D space via an adversarial process. Furthermore, we show that our approach can be viewed as a limit of existing notions of alignment by increasing transportation assignment tolerance. We quantitatively evaluate the method across domain adaptation tasks with shifts in label distributions. Our experiments show that the proposed method is more robust against these shifts than other alignment-based baselines.

📄 PDF Abstract BibTeX arXiv:2203.08908

Code (1)

timgaripov/asa 공식 구현 pytorch

Tasks

Domain Adaptation

Similar Papers 제목 키워드 기반

PGADA: Perturbation-Guided Adversarial Alignment for Few-shot Learning Under the Support-Query Shift

2022-05-08 · Siyang Jiang, Wei Ding, Hsi-Wen Chen, Ming-Syan Chen

Few-shot learning methods aim to embed the data to a low-dimensional embedding space and then classify the unseen query data to the seen support set. While these works assume that the support set and the query set lie in…

Data AugmentationFew-Shot Learning

Quantifying the Preferential Direction of the Model Gradient in Adversarial Training With Projected Gradient Descent

2020-09-10 · Ricardo Bigolin Lanfredi, Joyce D. Schroeder, Tolga Tasdizen

Adversarial training, especially projected gradient descent (PGD), has proven to be a successful approach for improving robustness against adversarial attacks. After adversarial training, gradients of models with respect…

Dual Adversarial Alignment for Realistic Support-Query Shift Few-shot Learning

2023-09-05 · Siyang Jiang, Rui Fang, Hsi-Wen Chen, Wei Ding 외

Support-query shift few-shot learning aims to classify unseen examples (query set) to labeled data (support set) based on the learned embedding in a low-dimensional space under a distribution shift between the support se…

Few-Shot Learning

SGM: A Framework for Building Specification-Guided Moderation Filters

2025-05-26 · Masoomali Fatehkia, Enes Altinisik, Husrev Taha Sencar

Aligning large language models (LLMs) with deployment-specific requirements is critical but inherently imperfect. Despite extensive training, models remain susceptible to misalignment and adversarial inputs such as jailb…

CASUAL: Conditional Support Alignment for Domain Adaptation with Label Shift

2023-05-29 · Anh T Nguyen, Lam Tran, Anh Tong, Tuan-Duy H. Nguyen 외

Unsupervised domain adaptation (UDA) refers to a domain adaptation framework in which a learning model is trained based on the labeled samples on the source domain and unlabeled ones in the target domain. The dominant ex…

Domain AdaptationUnsupervised Domain Adaptation