Research map / Reinforcement learning / Visual and representation RL
Suggesting Variable Order for Cylindrical Algebraic Decomposition via Reinforcement Learning
Fuqi Jia, Yuhang Dong, Minghao Liu, Pei Huang, Feifei Ma, Jian Zhang
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 unprojected-polynomial-free (NUP) instead of projected-polynomial (UP) input | GRL-SVO · 3-var test | #SI: 1772 -> 1798 | Improved |
| use GRL-SVO(UP) instead of gmods | expert heuristic · 9-var random | #SI: 286 -> 325 | Improved |
| use GRL-SVO(UP) instead of sotd on 4-var | expert heuristic · 4-var random | #SI: 625 -> 533 | Hurt |
| remove reward normalization factor M (ablation) | GRL-SVO(UP) · 3-var test | #SI: 1795 -> 1196 | Hurt |
| remove reward normalization factor M (ablation) | GRL-SVO(NUP) · 3-var test | #SI: 1769 -> 1666 | Hurt |
| remove the initial embedding (NO_EMB) (ablation) | GRL-SVO(NUP) · 3-var test | #SI: 1459 | Hurt |
| remove graph edges, using MLP (NO_EDGE) (ablation) | GRL-SVO(NUP) · 3-var test | #SI: 1763 | Mixed |
| use only one feature instead of the full feature set (ablation) | GRL-SVO · 3-var test | #SI: 1459 | Hurt |
| restrict to Sum operation features instead of Max (ablation) | GRL-SVO(UP) · 3-var test | #SI: 1716 -> 1788 | Improved |
| use degree features vs NoDegree features (UP) (ablation) | GRL-SVO(UP) · 3-var test | #SI: 1715 -> 1770 | Improved |
| replace GNN layer with GATConv (ablation) | GRL-SVO(NUP) · 3-var test | #SI: 1726 | Hurt |
| replace GNN layer with other architectures such as TransformerConv (ablation) | GRL-SVO(NUP) · 3-var test | #SI: 1769 | No clear effect |
Closest papers by meaning
- Automated Proof of Polynomial Inequalities via Reinforcement Learning (CVPR 2025 · similarity 0.85)
- Learning Polynomial Problems with $SL(2, \mathbb{R})$-Equivariance (ICLR 2024 · similarity 0.85)
- Computational Algebra with Attention: Transformer Oracles for Border Basis Algorithms (NeurIPS 2025 · similarity 0.84)
- Towards General Algorithm Discovery for Combinatorial Optimization: Learning Symbolic Branching Policy from Bipartite Graph (ICML 2024 · similarity 0.84)
- Learning to compute Gröbner bases (NeurIPS 2024 · similarity 0.83)
- Composing Global Solutions to Reasoning Tasks via Algebraic Objects in Neural Nets (NeurIPS 2025 · similarity 0.83)
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.
