All Stories

  1. Advanced Global Wildfire Activity Modeling with Hierarchical Graph ODE
  2. EmoRAG: Evaluating RAG Robustness to Symbolic Perturbations
  3. ACE: Concept Editing in Diffusion Models without Performance Degradation
  4. DynST: Dynamic Sparse Training for Resource-Constrained Spatio-Temporal Forecasting
  5. The Snowflake Hypothesis: Training and Powering GNN with One Node One Receptive Field
  6. The Heterophilic Snowflake Hypothesis: Training and Empowering GNNs for Heterophilic Graphs
  7. EXGC: Bridging Efficiency and Explainability in Graph Condensation
  8. Estimation of Urban Travel Time with Sparse Traffic Surveillance Data