Research map / Architectures and efficiency / RNNs and normalization
Abstraction based Output Range Analysis for Neural Networks
Pavithra Prabhakar, Zahra Rahimi Afzal
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 abstraction instead of encoding the full ACAS Xu network without abstraction | Output range analysis on ACAS Xu · ACAS Xu | MILP solving time | Improved |
| Increase the number of abstract nodes (ablation) | The proposed abstraction method · ACAS Xu | Output range upper bound | Improved |
| Increase the number of abstract nodes (ablation) | The proposed abstraction method · ACAS Xu | Computation time | Hurt |
| Change the partitioning of concrete neurons while keeping the number of abstract nodes fixed (ablation) | The proposed abstraction method · ACAS Xu | Output range upper bound: varies by a factor of 2 | Mixed |
Closest papers by meaning
- Exact Upper and Lower Bounds for the Output Distribution of Neural Networks with Random Inputs (ICML 2025 · similarity 0.82)
- Learning by Abstraction: The Neural State Machine (NeurIPS 2019 · similarity 0.81)
- Vector-output ReLU Neural Network Problems are Copositive Programs: Convex Analysis of Two Layer Networks and Polynomial-time Algorithms (ICLR 2021 · similarity 0.80)
- [Re] Nondeterminism and Instability in Neural Network Optimization (NeurIPS 2022 · similarity 0.79)
- The Separation Capacity of Random Neural Networks (NeurIPS 2023 · similarity 0.79)
- Variation Spaces for Multi-Output Neural Networks: Insights on Multi-Task Learning and Network Compression (NeurIPS 2024 · similarity 0.79)
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.
