paper-with-me

Papers

Autoencoding Under Normalization Constraints

2021-05-12 · Sangwoong Yoon, Yung-Kyun Noh, Frank Chongwoo Park

Likelihood is a standard estimate for outlier detection. The specific role of the normalization constraint is to ensure that the out-of-distribution (OOD) regime has a small likelihood when samples are learned using maximum likelihood. Because autoencoders do not possess such a process of normalization, they often fail to recognize outliers even when they are obviously OOD. We propose the Normalized Autoencoder (NAE), a normalized probabilistic model constructed from an autoencoder. The probability density of NAE is defined using the reconstruction error of an autoencoder, which is differently defined in the conventional energy-based model. In our model, normalization is enforced by suppressing the reconstruction of negative samples, significantly improving the outlier detection performance. Our experimental results confirm the efficacy of NAE, both in detecting outliers and in generating in-distribution samples.

📄 PDF Abstract BibTeX arXiv:2105.05735

Code (2)

swyoon/normalized-autoencoders 공식 구현 pytorch
heidelberg-hepml/normalized-autoencoders pytorch

Tasks

Outlier Detection

Similar Papers 제목 키워드 기반

Few-Shot and Zero-Shot Learning for Historical Text Normalization

2019-03-12 · WS 2019 11 · Marcel Bollmann, Natalia Korchagina, Anders Søgaard

Historical text normalization often relies on small training datasets. Recent work has shown that multi-task learning can lead to significant improvements by exploiting synergies with related datasets, but there has been…

LemmatizationMulti-Task LearningText NormalizationZero-Shot Learning

General Point Model with Autoencoding and Autoregressive

2023-10-25 · Zhe Li, Zhangyang Gao, Cheng Tan, Stan Z. Li 외

The pre-training architectures of large language models encompass various types, including autoencoding models, autoregressive models, and encoder-decoder models. We posit that any modality can potentially benefit from a…

DecoderLanguage ModelingLanguage ModellingLarge Language Model+3

General Point Model Pretraining with Autoencoding and Autoregressive

2024-01-01 · CVPR 2024 1 · Zhe Li, Zhangyang Gao, Cheng Tan, Bocheng Ren 외

The pre-training architectures of large language models encompass various types including autoencoding models autoregressive models and encoder-decoder models. We posit that any modality can potentially benefit from …

DecoderLanguage ModelingLanguage ModellingLarge Language Model+3

Evolution is Driven by Natural Autoencoding: Reframing Species, Interaction Codes, Cooperation, and Sexual Reproduction

2022-03-22 · Irun R. Cohen, Assaf Marron

The continuity of life and its evolution, we proposed, emerge from an interactive group process manifested in networks of interaction. We term this process \textit{survival-of-the-fitted}. Here, we reason that survival o…

Dimensionality Reduction

VAMAE: Vessel-Aware Masked Autoencoders for OCT Angiography

2026-04-08 · Ilerioluwakiiye Abolade, Prince Mireku, Kelechi Chibundu, Peace Ododo 외 arxiv

Optical coherence tomography angiography (OCTA) provides non-invasive visualization of retinal microvasculature, but learning robust representations remains challenging due to sparse vessel structures and strong topologi…

Self-Supervised Learning