Research map / Label-efficient and robust learning / Recognition and long-tail
Topic Modeling on Document Networks with Adjacent-Encoder
Read the paperCheck 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 |
|---|---|---|---|
| reconstruct neighbors with attention propagation (Adjacent-Encoder vs plain auto-encoder family) | auto-encoder topic model · Cora-DS transductive classification | accuracy: 0.558 -> 0.739 | Helped |
| add explicit adjacency vector input and reconstruction (AdjEnc-X) (ablation) | Adjacent-Encoder · Cora-ML inductive classification | accuracy: 0.761 -> 0.836 | Helped |
| add explicit adjacency vector input and reconstruction (AdjEnc-X) (ablation) | Adjacent-Encoder · Cora-ML transductive classification | accuracy: 0.864 -> 0.857 | No clear effect |
| add explicit adjacency vector (AdjEnc-X) (ablation) | Adjacent-Encoder · topic coherence PMI | PMI: DS 2.360 -> 1.872; ML 2.180 -> 2.337 | Mixed |
| add denoising (ablation) | Adjacent-Encoder · Cora-ML classification | accuracy: transductive 0.864 -> 0.855; inductive 0.761 -> 0.780 | Mixed |
| add denoising (ablation) | Adjacent-Encoder-X · Cora-ML classification | accuracy: 0.857 -> 0.865 transductive; 0.836 -> 0.839 inductive | Helped |
| add contractive regularization (ablation) | Adjacent-Encoder-X · Cora-ML classification | accuracy: 0.857 -> 0.872 transductive; 0.836 -> 0.844 inductive | Helped |
| add contractive regularization (ablation) | Adjacent-Encoder · Cora-ML classification | accuracy: 0.864 -> 0.856 transductive; 0.761 -> 0.780 inductive | Mixed |
| add K-sparse constraint (ablation) | Adjacent-Encoder · Cora-ML classification | accuracy: 0.864 -> 0.847 transductive; 0.761 -> 0.765 inductive | No clear effect |
| add K-sparse constraint (ablation) | Adjacent-Encoder-X · Cora-ML classification | accuracy: 0.857 -> 0.866 transductive; 0.836 -> 0.823 inductive | Mixed |
Closest papers by meaning
- Neural Topic Model via Optimal Transport (ICLR 2021 · similarity 0.87)
- Dynamic Topic Models for Temporal Document Networks (ICML 2022 · similarity 0.86)
- FASTopic: Pretrained Transformer is a Fast, Adaptive, Stable, and Transferable Topic Model (NeurIPS 2024 · similarity 0.85)
- Contextual Document Embeddings (ICLR 2025 · similarity 0.85)
- Hierarchical Retrieval: The Geometry and a Pretrain-Finetune Recipe (NeurIPS 2025 · similarity 0.83)
- Repurposing Language Models into Embedding Models: Finding the Compute-Optimal Recipe (NeurIPS 2024 · similarity 0.83)
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.