{"total":16,"items":[{"citing_arxiv_id":"2605.19916","ref_index":31,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Fast and Featureless Node Representation Learning with Partial Pairwise Supervision","primary_cat":"cs.LG","submitted_at":"2026-05-19T14:40:51+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"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.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18579","ref_index":22,"ref_count":2,"confidence":0.9,"is_internal_anchor":false,"paper_title":"S2Aligner: Pair-Efficient and Transferable Pre-Training for Sparse Text-Attributed Graphs","primary_cat":"cs.LG","submitted_at":"2026-05-18T15:56:19+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"S2Aligner decouples semantic and structural components in LLM-as-Aligner pre-training for sparse TAGs and uses structure-oriented reconstruction plus domain risk balancing to improve transferability and reduce generalization gaps.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.20248","ref_index":29,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Graph Transductive Sharpening: Leveraging Unlabeled Predictions in Node Classification","primary_cat":"cs.LG","submitted_at":"2026-05-18T06:47:41+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Transductive Sharpening adds an entropy-minimization term on unlabeled-node predictions to the training objective for graph node classification.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.09862","ref_index":41,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"UFO: A Unified Flow-Oriented Framework for Robust Continual Graph Learning","primary_cat":"cs.LG","submitted_at":"2026-05-11T01:49:05+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"UFO combines flow-based generative replay with instance-level reliability scoring to handle both catastrophic forgetting and catastrophic remembering from noisy supervision in evolving graphs, outperforming baselines on four datasets.","context_count":1,"top_context_role":"dataset","top_context_polarity":"use_dataset","context_text":"For the current task, the final objective for graph knowledge preservation is: LKP = αELE + αLLL, (13) where αE and αL are loss weights. The whole training process is summarized in Algorithm 1. 5 Experiments 5.1 Experimental Setup Datasets and Implementation Details. We evaluate our method on four real-world datasets: Cora- Full [ 2], CS [ 40], WikiCS [ 41], and Photo [ 40]. More detailed statistics are provided in Ap- pendix A.1. Following [ 42], we divide the classes into a sequence of tasks for each dataset, and split the nodes in each task into 60%, 20%, and 20% for training, validation, and testing, respec- tively. We add symmetric noise to a fraction of labels in the training set while keeping the validation"},{"citing_arxiv_id":"2605.03514","ref_index":13,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Revisiting Graph-Tokenizing Large Language Models: A Systematic Evaluation of Graph Token Understanding","primary_cat":"cs.CL","submitted_at":"2026-05-05T08:50:28+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"GTokenLLMs do not fully understand graph tokens, exhibiting over-sensitivity or insensitivity to instruction changes and relying heavily on text for reasoning even when graph information is preserved.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.27462","ref_index":2,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Improving Graph Few-shot Learning with Hyperbolic Space and Denoising Diffusion","primary_cat":"cs.LG","submitted_at":"2026-04-30T06:02:16+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"IMPRESS improves graph few-shot learning by learning representations in hyperbolic space and using denoising diffusion to better approximate target distributions from few support samples.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.17411","ref_index":10,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"DuConTE: Dual-Granularity Text Encoder with Topology-Constrained Attention for Text-attributed Graphs","primary_cat":"cs.CL","submitted_at":"2026-04-19T12:36:18+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"DuConTE is a dual-granularity text encoder that incorporates graph topology into language model attention for improved node representations in text-attributed graphs.","context_count":1,"top_context_role":"method","top_context_polarity":"use_method","context_text":"we allow attention only between node representations that correspond to the same node or to connected nodes in the graph. Accordingly, the attention mask matrix M node mask is constructed as follows: for any two positions m and n in Z (i), let v(m) and v(n) denote the corresponding nodes in the graph. The entry M node m,n ∈ {0,1} (k+1)×(k+1) is defined as: M node m,n = \u001a1ifv(m) =v(n)or(v(m), v(n))∈ E, 0otherwise. (10) 4.3 Node Representation Composer To effectively fuse the word-token embeddingsH (i) into high-quality first-stage node representations, we design a Node Representation Composer f. The composer employs two distinct modules: a more sophisticated module f1 to compute the representation of the target nodevi, and a lightweight module f2 to independently encode each neighbor nodevj ∈ N(i) ."},{"citing_arxiv_id":"2604.14746","ref_index":27,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Disentangle-then-Refine: LLM-Guided Decoupling and Structure-Aware Refinement for Graph Contrastive Learning","primary_cat":"cs.AI","submitted_at":"2026-04-16T07:57:11+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"SDM-SCR uses LLMs for semantic disentanglement of signal from noise in text-attributed graphs followed by spectral consistency regularization to improve contrastive learning performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.08980","ref_index":18,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Neighbourhood Transformer: Switchable Attention for Monophily-Aware Graph Learning","primary_cat":"cs.LG","submitted_at":"2026-04-10T05:35:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Neighbourhood Transformers apply local self-attention for monophily-aware graph learning, guarantee expressiveness at least as strong as message-passing GNNs, and outperform prior methods on node classification across ten datasets while cutting memory and time costs substantially.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.06391","ref_index":34,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Toward a universal foundation model for graph-structured data","primary_cat":"cs.LG","submitted_at":"2026-04-07T19:21:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A pretrained graph model using feature-agnostic structural prompts matches or exceeds supervised baselines and shows strong zero-shot and few-shot transfer on held-out biomedical graphs, with a 21.8% ROC-AUC gain on SagePPI.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2603.02938","ref_index":21,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Beyond One-Size-Fits-All: Adaptive Subgraph Denoising for Zero-Shot Graph Learning with Large Language Models","primary_cat":"cs.LG","submitted_at":"2026-03-03T12:47:44+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"GraphSSR introduces an adaptive SSR pipeline with SSR-SFT data synthesis and SSR-RL (Authenticity-Reinforced and Denoising-Reinforced stages) to overcome one-size-fits-all subgraph noise in zero-shot LLM graph reasoning.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2602.17071","ref_index":77,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"AdvSynGNN: Structure-Adaptive Graph Neural Nets via Adversarial Synthesis and Self-Corrective Propagation","primary_cat":"cs.LG","submitted_at":"2026-02-19T04:26:57+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"AdvSynGNN uses multi-resolution structural synthesis, contrastive objectives, an adaptive transformer, and an adversarial propagation engine with residual label correction to improve node-level predictions on challenging graph topologies.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2602.11629","ref_index":4,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"GP2F: Cross-Domain Graph Prompting with Adaptive Fusion of Pre-trained Graph Neural Networks","primary_cat":"cs.LG","submitted_at":"2026-02-12T06:25:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"GP2F is a dual-branch graph prompting framework that fuses frozen pre-trained knowledge with task-specific adaptation to reduce estimation error and outperform baselines in cross-domain few-shot node and graph classification.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2512.24062","ref_index":29,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Energy-Balanced Hyperspherical Graph Representation Learning via Structural Binding and Entropic Dispersion","primary_cat":"cs.LG","submitted_at":"2025-12-30T08:11:37+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"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.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2510.08952","ref_index":32,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"When LLM Agents Meet Graph Optimization: An Automated Data Quality Improvement Approach","primary_cat":"cs.LG","submitted_at":"2025-10-10T02:59:19+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"LAGA is a unified multi-agent LLM framework that automates comprehensive quality optimization for text-attributed graphs by running detection, planning, action, and evaluation agents in a closed loop.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2508.07117","ref_index":3,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"From Nodes to Narratives: Explaining Graph Neural Networks with LLMs and Graph Context","primary_cat":"cs.LG","submitted_at":"2025-08-09T23:22:38+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"GSPELL projects GNN embeddings into LLM space and builds hybrid prompts to produce faithful natural-language explanations and sparse subgraphs for GNN predictions on text-attributed graphs.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}