Research map / Probabilistic and scientific modeling / Variational and generative sampling
Latent Space Autoregression for Novelty Detection
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Combine reconstruction error and latent log-likelihood in the novelty score (ablation) | Latent Space Autoregression · MNIST | AUROC: 0.949 -> 0.975 | Helped |
| Combine reconstruction error and latent log-likelihood in the novelty score (ablation) | Latent Space Autoregression · CIFAR-10 | AUROC: 0.603 -> 0.641 | Helped |
| Combine reconstruction error and latent log-likelihood in the novelty score (ablation) | Latent Space Autoregression · UCSD Ped2 | AUROC: 0.909 -> 0.954 | Helped |
| Combine reconstruction error and latent log-likelihood in the novelty score (ablation) | Latent Space Autoregression · ShanghaiTech | AUROC: 0.726 -> 0.725 | Mixed |
| Combine reconstruction error and latent log-likelihood in the novelty score (ablation) | Latent Space Autoregression · DR(eye)VE | AUROC: 0.863 -> 0.926 | Helped |
| Use CIFAR-10-specific supervised semantic features instead of raw images (ablation) | Latent Space Autoregression · CIFAR-10 | AUROC: 0.64 -> 0.99 | Helped |
| Use ImageNet-pretrained semantic features instead of raw images (ablation) | Latent Space Autoregression · CIFAR-10 | AUROC: 0.64 -> 0.72 | Helped |
| Use a flexible autoregressive latent prior rather than a fixed isotropic Gaussian prior | Latent Space Autoregression · MNIST | AUROC: VAE 0.969; ours 0.975 | Helped |
| Use a flexible autoregressive latent prior rather than a fixed isotropic Gaussian prior | Latent Space Autoregression · CIFAR-10 | AUROC: VAE 0.586; ours 0.641 | Helped |
| Use the proposed MFC/MSC autoregressive layers instead of recurrent LSTM layers (ablation) | Latent Space Autoregression · CIFAR-10 and UCSD Ped2 | AUROC | Helped |
Closest papers by meaning
- Generative Probabilistic Novelty Detection with Adversarial Autoencoders (NeurIPS 2018 · similarity 0.88)
- Universal Novelty Detection Through Adaptive Contrastive Learning (CVPR 2024 · similarity 0.87)
- AutoSciDACT: Automated Scientific Discovery through Contrastive Embedding and Hypothesis Testing (NeurIPS 2025 · similarity 0.86)
- ALTBI: Constructing Improved Outlier Detection Models via Optimization of Inlier-Memorization Effect (AAAI 2025 · similarity 0.85)
- Beyond Normal References: Discriminative Few-Shot Anomaly Detection (NeurIPS 2026 · similarity 0.85)
- VACE: Learning Geometrically Structured Representations for Time Series Anomaly Detection (NeurIPS 2026 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.