Research map / Graph learning / Graph algorithms and clustering
Neuron Dependency Graphs: A Causal Abstraction of Neural Networks
Yaojie Hu, Jin Tian
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 |
|---|---|---|---|
| Use random-initialized (untrained) model instead of trained model for NDG extraction (ablation) | NDG edges · AllNLI, CUB200, Devign | edge accuracy on test set: test accuracy ~50% vs ~99% for trained | Hurt |
| Feed random inputs to trained model instead of real inputs (ablation) | NDG edges · MNIST, SST2, AllNLI, CUB200, Devign | edge test accuracy: e.g. CUB200 91.80 vs 97.15 original; AllNLI 99.55 vs 99.58 | Mixed |
| Build inter-model edges between two independently trained models on same input (ablation) | NDG · MNIST even, SST2, AllNLI, Devign | edge count and test accuracy: 0 edges on MNIST and CUB200; 35354 edges on SST2 with 92.29 test accuracy | Mixed |
| Same model with two disjoint input halves (sanity check) (ablation) | NDG · all datasets | number of inter-model edges: 0 edges on all datasets | No clear effect |
| Increase threshold alpha from low to high values (ablation) | NDG edges · all datasets | training and test edge accuracy: accuracy generally increases with alpha | Improved |
| Interchange intervention on NDG-aligned neurons (Algorithm 1) | neural network predictions · MNIST, MNIST even, SST2, AllNLI, CUB200, Devign | aligned percentage on test set: 88.39% to 100% aligned | Improved |
Closest papers by meaning
- Rethinking Causal Relationships Learning in Graph Neural Networks (AAAI 2024 · similarity 0.86)
- Neural Causal Abstractions (AAAI 2024 · similarity 0.86)
- Towards Learning and Explaining Indirect Causal Effects in Neural Networks (AAAI 2024 · similarity 0.86)
- Generating Hypotheses of Dynamic Causal Graphs in Neuroscience: Leveraging Generative Factor Models of Observed Time Series (ICML 2025 · similarity 0.86)
- Causal Concept Graph Models: Beyond Causal Opacity in Deep Learning (ICLR 2025 · similarity 0.85)
- Causal pieces: analysing and improving spiking neural networks piece by piece (NeurIPS 2026 · similarity 0.85)
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.
