Research map / Probabilistic and scientific modeling / Treatment effects
Learning Exposure Mapping Functions for Inferring Heterogeneous Peer Effects
Shishir Adhikari, Sourav Medya, Elena Zheleva
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| learn exposure mapping with EgoNetGnn (CFR head) vs best baseline | GNN-TARNet-Motifs · BlogCatalog, clustering coefficient | PEHE error: 2.04 -> 1.51 | Improved |
| EgoNetGnn-CFR vs INE-TARNet | INE-TARNet · BlogCatalog, mutual connections | PEHE error: 3.37 -> 2.57 | Improved |
| EgoNetGnn-CFR vs CauGramer | CauGramer · BlogCatalog, attribute similarity | PEHE error: 7.34 -> 4.86 | Improved |
| use CFR (balance loss) instead of TARNet head | EgoNetGnn · BlogCatalog, clustering coefficient | PEHE error: 1.59 -> 1.51 | Improved |
| add masked weights (ablation) | EgoNetGnn-TARNet · BA, attribute similarity | PEHE error: 2.37 -> 1.23 | Improved |
| add masked weights (ablation) | EgoNetGnn-TARNet · BC, mutual connections | PEHE error: 2.91 -> 2.90 | No clear effect |
| add masked weights (ablation) | EgoNetGnn-TARNet · BC, clustering coefficient | PEHE error: 1.98 -> 1.59 | Improved |
| add feature encoder (ablation) | EgoNetGnn-TARNet · BA, attribute similarity | PEHE error: 13.73 -> 1.23 | Improved |
| add feature encoder (ablation) | EgoNetGnn-TARNet · BC, clustering coefficient | PEHE error: 1.55 -> 1.59 | Hurt |
| add feature encoder (ablation) | EgoNetGnn-TARNet · WS, clustering coefficient | PEHE error: 1.91 -> 1.18 | Improved |
| learn exposure vs fraction of treated friends | EgoNetGnn · BlogCatalog, clustering coefficient | correlation with true exposure: 0.17 -> 0.81 | Improved |
| learn exposure vs fraction of treated friends | EgoNetGnn · BlogCatalog, attribute similarity | correlation with true exposure: 0.61 -> 0.50 | Hurt |
Closest papers by meaning
- Telling Peer Direct Effects from Indirect Effects in Observational Network Data (ICML 2025 · similarity 0.89)
- Using Embeddings for Causal Estimation of Peer Influence in Social Networks (NeurIPS 2022 · similarity 0.89)
- Heterogeneous Peer Effects in the Linear Threshold Model (AAAI 2022 · similarity 0.87)
- Embedding Network Autoregression for Time Series Analysis and Causal Peer Effect Inference (JMLR 2026 · similarity 0.87)
- Online Experimental Design With Estimation-Regret Trade-off Under Network Interference (NeurIPS 2025 · similarity 0.83)
- Influence Functions for Edge Edits in Non-Convex Graph Neural Networks (NeurIPS 2025 · similarity 0.82)
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.
