Cellina uses supervised disentanglement to separate cell intrinsic states from spatial contexts for counterfactual predictions on tissue graphs, outperforming baselines on 2.5M+ cells from cancer and brain data.
hub
Variational Graph Auto-Encoders
51 Pith papers cite this work, alongside 897 external citations. Polarity classification is still indexing.
51
Pith papers citing it
897
external citations · Pith
abstract
We introduce the variational graph auto-encoder (VGAE), a framework for unsupervised learning on graph-structured data based on the variational auto-encoder (VAE). This model makes use of latent variables and is capable of learning interpretable latent representations for undirected graphs. We demonstrate this model using a graph convolutional network (GCN) encoder and a simple inner product decoder. Our model achieves competitive results on a link prediction task in citation networks. In contrast to most existing models for unsupervised learning on graph-structured data and link prediction, our model can naturally incorporate node features, which significantly improves predictive performance on a number of benchmark datasets.
hub tools
citation-role summary
background 4
citation-polarity summary
fields
cs.LG 30 cs.SI 5 cs.CL 4 cs.CV 3 cond-mat.mtrl-sci 1 cs.CE 1 cs.CR 1 cs.IR 1 math.NA 1 q-bio.GN 1roles
background 4representative citing papers
GRAIL trains graph predictors via imitation learning by modeling generation as sequential decisions on partial graph embeddings, matching or exceeding prior methods on 18 benchmarks.
HyperGRL places graph nodes on a hypersphere and minimizes Helmholtz free energy with structural binding energy and mean-field repulsive potential, regulated by an adaptive thermostat, to produce discriminative representations.
GRDM jointly generates relational database tables via graph-conditional diffusion without table ordering, outperforming autoregressive baselines on multi-hop correlations and single-table fidelity across six real RDBs.
A multi-task scheme with synthetic anomalies from graph perturbations and two-phase training learns robust features for weakly supervised graph anomaly detection, outperforming competitors on public datasets.
Feature reconstruction in GSSL is robust to noise in text-driven biomedical graphs while relation reconstruction is sensitive, with bidirectional GNN architectures performing better on noisy data and yielding up to 7% gains over language model baselines.
A structure-aware VAE generates realistic FC matrices for replay, combined with multi-level knowledge distillation and hierarchical contextual bandit sampling, to enable continual fMRI-based brain disorder diagnosis across sequentially arriving multi-site data without catastrophic forgetting.
SAOT applies structure-aware optimal transport to capture global inter-node correspondences and uses cross-task distillation to retain prior structural knowledge, yielding accuracy gains of up to 15% on Products-CL in class-incremental settings.
FedLAB organizes multimodal graph knowledge into typed hierarchical codebooks for modality evidence, node semantics, and topology context via federated semantic barycenter pre-training, improving performance by up to 7.53% on benchmarks while enabling semantic traceability.
Dynamic Bayesian predictive synthesis combines forecasts from multiple network mechanisms with time-varying weights for adaptive edge prediction and mechanism identification in dynamic networks.
OPD-Evolver uses on-policy self-distillation in fast interaction and slow attribution loops to build agents with holistic memory competence, outperforming prior systems by up to 11.5% and allowing a 9B model to compete with much larger ones.
FLAGG makes one-shot graph generators autoregressive through a learned reversal of stochastic node removal and outperforms baselines on datasets spanning different graph sizes and domains.
Graph Cascades uses contagion diffusion to rewire graphs by promoting reinforced multi-hop node pairs to direct neighbors, improving GNN performance on heterophilic and moderate-degree homophilic graphs under specified conditions.
GPR-GAE is a novel self-supervised graph auto-encoder using multiple Generalized PageRank filters that serves as a plug-and-play purifier achieving state-of-the-art robustness for GNNs against structural attacks.
TopoPrior learns transferable topology priors offline from multi-domain reference graphs using a conditional variational graph model and adversarial adaptation to initialize collaboration structures for multi-agent LLM systems, reducing online search overhead.
GraViti introduces a graph-level VAE with relaxed permutation invariance that maps whole graphs to latent vectors, achieves strong reconstruction on large molecular datasets, and generates valid samples by learning constraints directly from graph-level representations.
DART mitigates structural overfitting in graph missing-feature imputation via global structural augmentation, masked-autoencoder semantic rectification, and test-time distribution rectification, outperforming prior methods on transductive and inductive tasks including a new real-missing dataset.
GraphBSI uses Bayesian Sample Inference as noise-controlled SDEs to generate discrete graphs in one shot, achieving state-of-the-art results on molecular benchmarks Moses and GuacaMol.
GEMS formulates close-ended human-behavior simulation as link prediction on a heterogeneous graph and matches or exceeds LLM performance with three orders of magnitude fewer parameters across three datasets and three evaluation settings.
SSL4RL reformulates self-supervised learning objectives into dense, verifiable reward signals for RL-based fine-tuning of vision-language models, yielding performance gains on reasoning benchmarks.
GABI learns geometry-conditioned latent priors from multi-geometry physical response datasets for use in Bayesian inversion, yielding geometry-adapted posteriors via ABC sampling.
The paper proposes the Heterogeneous Temporal HyperGraph Neural Network (HTHGN) with a new hyperedge construction algorithm, hierarchical attention for temporal message passing, and contrastive learning to model high-order interactions in heterogeneous temporal graphs, reporting performance gains on
Entity representations learned from text via link prediction generalize to unseen entities and transfer to classification and retrieval with reported gains of 22% MRR, 16% accuracy, and 8.8% NDCG@10.
ML-GCN embeds nodes via GCN, generates a label matrix in the same space, and trains with relaxed skip-gram on node-label concatenations to model correlations, reporting outperformance on graph datasets.
citing papers explorer
-
Querying Counterfactuals on Tissue Graphs with Supervised Disentanglement
Cellina uses supervised disentanglement to separate cell intrinsic states from spatial contexts for counterfactual predictions on tissue graphs, outperforming baselines on 2.5M+ cells from cancer and brain data.
-
Building Deep Graph Predictors with Graph Imitation Learning
GRAIL trains graph predictors via imitation learning by modeling generation as sequential decisions on partial graph embeddings, matching or exceeding prior methods on 18 benchmarks.
-
Energy-Balanced Hyperspherical Graph Representation Learning via Structural Binding and Entropic Dispersion
HyperGRL places graph nodes on a hypersphere and minimizes Helmholtz free energy with structural binding energy and mean-field repulsive potential, regulated by an adaptive thermostat, to produce discriminative representations.
-
Joint Relational Database Generation via Graph-Conditional Diffusion Models
GRDM jointly generates relational database tables via graph-conditional diffusion without table ordering, outperforming autoregressive baselines on multi-hop correlations and single-table fidelity across six real RDBs.
-
Learning Feature Encoder with Synthetic Anomalies for Weakly Supervised Graph Anomaly Detection
A multi-task scheme with synthetic anomalies from graph perturbations and two-phase training learns robust features for weakly supervised graph anomaly detection, outperforming competitors on public datasets.
-
Robustness of Graph Self-Supervised Learning to Real-World Noise: A Case Study on Text-Driven Biomedical Graphs
Feature reconstruction in GSSL is robust to noise in text-driven biomedical graphs while relation reconstruction is sensitive, with bidirectional GNN architectures performing better on noisy data and yielding up to 7% gains over language model baselines.
-
Continual Learning for fMRI-Based Brain Disorder Diagnosis via Functional Connectivity Matrices Generative Replay
A structure-aware VAE generates realistic FC matrices for replay, combined with multi-level knowledge distillation and hierarchical contextual bandit sampling, to enable continual fMRI-based brain disorder diagnosis across sequentially arriving multi-site data without catastrophic forgetting.
-
SAOT: Self-Supervised Continual Graph Learning with Structure-Aware Optimal Transport
SAOT applies structure-aware optimal transport to capture global inter-node correspondences and uses cross-task distillation to retain prior structural knowledge, yielding accuracy gains of up to 15% on Products-CL in class-incremental settings.
-
FedLAB: Traceable Semantic Codebooks for Federated Multimodal Graph Foundation Learning
FedLAB organizes multimodal graph knowledge into typed hierarchical codebooks for modality evidence, node semantics, and topology context via federated semantic barycenter pre-training, improving performance by up to 7.53% on benchmarks while enabling semantic traceability.
-
Bayesian Predictive Synthesis for Dynamic Networks: Forecasting and Identifying Structural Mechanisms
Dynamic Bayesian predictive synthesis combines forecasts from multiple network mechanisms with time-varying weights for adaptive edge prediction and mechanism identification in dynamic networks.
-
OPD-Evolver: Cultivating Holistic Agent Evolver via On-Policy Distillation
OPD-Evolver uses on-policy self-distillation in fast interaction and slow attribution loops to build agents with holistic memory competence, outperforming prior systems by up to 11.5% and allowing a 9B model to compete with much larger ones.
-
FLAGG: Flexible Autoregressive Graph Generation
FLAGG makes one-shot graph generators autoregressive through a learned reversal of stochastic node removal and outperforms baselines on datasets spanning different graph sizes and domains.
-
Graph Cascades: Contagion-Based Mesoscopic Rewiring for Structure-Aware Graph Machine Learning
Graph Cascades uses contagion diffusion to rewire graphs by promoting reinforced multi-hop node pairs to direct neighbors, improving GNN performance on heterophilic and moderate-degree homophilic graphs under specified conditions.
-
Self-supervised Adversarial Purification for Graph Neural Networks
GPR-GAE is a novel self-supervised graph auto-encoder using multiple Generalized PageRank filters that serves as a plug-and-play purifier achieving state-of-the-art robustness for GNNs against structural attacks.
-
Learning Transferable Topology Priors for Multi-Agent LLM Collaboration Across Domains
TopoPrior learns transferable topology priors offline from multi-domain reference graphs using a conditional variational graph model and adversarial adaptation to initialize collaboration structures for multi-agent LLM systems, reducing online search overhead.
-
GraViti: Graph-Level Variational Autoencoders with Relaxed Permutation Invariance
GraViti introduces a graph-level VAE with relaxed permutation invariance that maps whole graphs to latent vectors, achieves strong reconstruction on large molecular datasets, and generates valid samples by learning constraints directly from graph-level representations.
-
Mitigating Structural Overfitting: A Distribution-Aware Rectification Framework for Missing Feature Imputation
DART mitigates structural overfitting in graph missing-feature imputation via global structural augmentation, masked-autoencoder semantic rectification, and test-time distribution rectification, outperforming prior methods on transductive and inductive tasks including a new real-missing dataset.
-
Discrete Bayesian Sample Inference for Graph Generation
GraphBSI uses Bayesian Sample Inference as noise-controlled SDEs to generate discrete graphs in one shot, achieving state-of-the-art results on molecular benchmarks Moses and GuacaMol.
-
Graph-Based Alternatives to LLMs for Human Simulation
GEMS formulates close-ended human-behavior simulation as link prediction on a heterogeneous graph and matches or exceeds LLM performance with three orders of magnitude fewer parameters across three datasets and three evaluation settings.
-
SSL4RL: Revisiting Self-supervised Learning as Intrinsic Reward for Visual-Language Reasoning
SSL4RL reformulates self-supervised learning objectives into dense, verifiable reward signals for RL-based fine-tuning of vision-language models, yielding performance gains on reasoning benchmarks.
-
Geometric Autoencoder Priors for Bayesian Inversion: Learn First Observe Later
GABI learns geometry-conditioned latent priors from multi-geometry physical response datasets for use in Bayesian inversion, yielding geometry-adapted posteriors via ABC sampling.
-
Heterogeneous Temporal Hypergraph Neural Network
The paper proposes the Heterogeneous Temporal HyperGraph Neural Network (HTHGN) with a new hyperedge construction algorithm, hierarchical attention for temporal message passing, and contrastive learning to model high-order interactions in heterogeneous temporal graphs, reporting performance gains on
-
Inductive Entity Representations from Text via Link Prediction
Entity representations learned from text via link prediction generalize to unseen entities and transfer to classification and retrieval with reported gains of 22% MRR, 16% accuracy, and 8.8% NDCG@10.
-
Semi-Supervised Graph Embedding for Multi-Label Graph Node Classification
ML-GCN embeds nodes via GCN, generates a label matrix in the same space, and trains with relaxed skip-gram on node-label concatenations to model correlations, reporting outperformance on graph datasets.
-
A Unified Benchmark for Evaluating Knowledge Graph Construction Methods and Graph Neural Networks
A dual-purpose benchmark supplies two text-derived knowledge graphs and one expert reference graph on the same biomedical corpus to jointly measure construction method quality and GNN robustness via semi-supervised node classification.
-
Graph Federated Unlearning for Privacy Preservation
Orthogonal unlearning updates plus server-side virtual clients enable effective user data removal in graph federated learning without major performance loss.
-
H3: A Healthcare Three-Hop Index for Physician Referral Network Prediction
H3 is a new three-hop index that predicts physician referrals using normalized indirect pathways and outperforms heuristics and neural nets on Medicare shared-patient data in both within-period and cross-period settings.
-
Mochi: Aligning Pre-training and Inference for Efficient Graph Foundation Models via Meta-Learning
Mochi aligns pre-training with inference via meta-learning for efficient graph foundation models, matching or exceeding prior models on 25 datasets with 8-27x less training time.
-
Spectral Embeddings Leak Graph Topology: Theory, Benchmark, and Adaptive Reconstruction
LoGraB creates fragmented graph benchmarks with controls for radius, spectral quality, noise, and coverage, while AFR reconstructs faithful graph islands from spectral patches using fidelity scoring, RANSAC-Procrustes alignment, and adaptive stitching, supported by recovery proofs and strong results
-
A Physics-Aware Variational Graph Autoencoder for Joint Modal Identification with Uncertainty Quantification
A physics-informed graph variational autoencoder jointly predicts modal frequencies, damping, and shapes from PSD data of trusses with uncertainty quantification and orthogonality constraints.
-
Multi-Label Node Classification with Label Influence Propagation
LIP decomposes GNN message passing to quantify label influences, builds a label influence graph, and propagates high-order effects to outperform prior methods on multi-label node classification benchmarks.
-
Estimation-Prediction Tradeoff in Causal Probabilistic Temporal Graphs
In binary logistic temporal-graph models, higher Fisher information for parameter recovery coincides with higher irreducible predictive entropy, so the easiest-to-estimate regimes are the hardest to predict.
-
Fast and Featureless Node Representation Learning with Partial Pairwise Supervision
Contrastive FUSE learns node embeddings from partial pairwise supervision and structural signals alone by optimizing a spectral contrastive objective with a lightweight modularity approximation, yielding competitive performance and runtime gains on citation and co-purchase graphs.
-
Decoder-only Clustering in Attributed Graphs
A new method for nodal clustering in attributed graphs that combines low-dimensional node priors, a neural decoder, and graph-fused LASSO regularization on prior means, demonstrated via simulations on grid graphs and real data applications.
-
Disentangled Generative Graph Representation Learning
DiGGR introduces a self-supervised graph representation learning framework that disentangles latent factors to guide mask modeling and improve representation quality on graph tasks.
-
Explaining Graph Neural Networks for Node Similarity on Graphs
Empirical comparison shows gradient-based explanations for GNN node similarities are actionable, consistent, and retain effects when sparsified, unlike mutual information explanations.
-
$\texttt{SynC}$: Synergistic Boosting of Structure and Representation for Deep Graph Clustering
SynC introduces a shared-weight TIGAE-based framework that alternates embedding learning and structure augmentation for graph clustering, with added fine-tuning for low-homophily graphs and reported gains on benchmarks.
-
CROSS-Net: Region-Agnostic Taxi-Demand Prediction Using Feature Disentanglement
CROSS-Net combines multiview graph neural networks with VAE-based feature disentanglement to predict taxi demand in previously unseen urban regions.
-
Spectral-based Graph Convolutional Network for Directed Graphs
A spectral-based GCN for directed graphs uses redefined Laplacians to enable direct application to directed data and outperforms prior methods on semi-supervised node classification tasks.
-
Astro Generative Network: A Variational Framework for Controlled Node Insertion in Incomplete Complex Networks
AGN is a variational framework for inserting plausible new nodes into incomplete networks by latent sampling and similarity attachment, shown on synthetic data to keep clustering and modularity changes modest compared to a baseline that allows new-new edges.
-
Combining opinion and structural similarity in link recommendations to counter extreme polarization
Weak structural similarity combined with strong opinion similarity in link recommendations prevents network fragmentation and favors moderate opinions under strong homophily.
-
Frequency-Corrupt Based Graph Self-Supervised Learning
FC-GSSL is a graph self-supervised method that corrupts nodes/edges with high low-frequency contribution and reconstructs low-frequency/general targets, improving node and graph prediction on most tested benchmarks.
-
From Load Tests to Live Streams: Graph Embedding-Based Anomaly Detection in Microservice Architectures
A GCN-GAE model learns node embeddings from directed weighted microservice graphs to flag anomalies via cosine similarity between load-test and live-event representations, with a synthetic injection framework reporting 96% precision.
-
Extracting Money Laundering Transactions from Quasi-Temporal Graph Representation
ExSTraQt uses quasi-temporal graph representations and supervised learning to detect suspicious transactions, achieving F1 score uplifts of up to 1% on real data and over 8% on synthetic datasets compared to prior AML models.
-
PolyGraphPy: A unified Python framework for atomistic simulation and machine learning-driven polymer design
A unified Python framework integrates DFTB+ simulation, Bayesian GNNs, and GPT/GA generators to predict and design acrylate polymers with targeted polarizability, validated only on internal DFTB-derived data.
-
AutoGraphAD: Unsupervised network anomaly detection using Variational Graph Autoencoders
A heterogeneous VGAE with masking, edge dropping, and negative sampling flags anomalous network connections via a weighted reconstruction/regularization score, matching Anomal-E F1 with fewer false positives but lower recall, and faster inference.
-
A Hierarchical Quantized Tokenization Framework for Task-Adaptive Graph Representation Learning
QUIET is a hierarchical RVQ-based graph tokenizer with a learned level-weighting gate; it improves several benchmarks but not consistently against the strongest baselines.
-
Leveraging Medical Foundation Model Features in Graph Neural Network-Based Retrieval of Breast Histopathology Images
A graph autoencoder model using foundation model features achieves high retrieval accuracy (mAP 96.7-97.6%, mMV 91.5-94.2%) on BreakHis and BACH breast cancer histopathology datasets.
-
Graph Embeddings at Scale
Presents a distributed infrastructure for scaling skip-gram graph embeddings to 68M-vertex networks by avoiding partitioning, using dynamic size-constrained graphs, and efficient indexing for updates.
-
Detecting Complex Money Laundering Patterns with Incremental and Distributed Graph Modeling
ReDiRect uses fuzzy graph partitioning for distributed unsupervised money laundering detection and a refined evaluation metric, claiming better efficiency than prior methods on Libra and IBM datasets.