Research map / Learning theory and algorithms / Combinatorial and search optimization
Interpreting Neural Network Judgments via Minimal, Stable, and Symbolic Corrections
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Raise the maximum number of explored regions (ablation) | Judgment interpretation algorithm · Mortgage underwriting, four selected inputs | correction volume: 2.4, 10.3, 9.2, 1.29 -> 2.4, 26.3, 21.9, 11.2 | Helped |
| Raise the maximum number of explored regions (ablation) | Judgment interpretation algorithm · Mortgage underwriting, four selected inputs | time (in seconds): 102, 191, 141, 118 -> 101, 375, 291, 1655 | Hurt |
Closest papers by meaning
- Sound and Complete Neural Network Repair with Minimality and Locality Guarantees (ICLR 2022 · similarity 0.85)
- Verification of Non-Linear Specifications for Neural Networks (ICLR 2019 · similarity 0.84)
- Dynamic Back-Substitution in Bound-Propagation-Based Neural Network Verification (AAAI 2025 · similarity 0.83)
- Verification Learning: Make Unsupervised Neuro-Symbolic System Feasible (ICML 2025 · similarity 0.83)
- Explaining, Fast and Slow: Abstraction and Refinement of Provable Explanations (ICML 2025 · similarity 0.83)
- Mastering Symbolic Operations: Augmenting Language Models with Compiled Neural Networks (ICLR 2024 · 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.