Research map / Probabilistic and scientific modeling / Variational and generative sampling
GenSR: Symbolic regression based on equation generative space
Qian Li, Yuxiao Hu, Juncheng Liu, Yuntian Chen
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 the posterior branch instead of the prior branch for reconstruction (ablation) | GenSR CVAE · Synthetic LogExp and Trig datasets | Levenshtein distance: 0.106 -> 0.094 (LogExp-15); 0.171 -> 0.146 (Trig-25) | Improved |
| increase expression length and add trigonometric operators | GenSR CVAE · Synthetic reconstruction | Levenshtein distance: 0.094 -> 0.146 (posterior, LogExp-15 -> Trig-25) | Hurt |
| increase latent dimension from 64 to 512 (ablation) | GenSR pretrained model (no CMA-ES) · SRBench noise-free | R2 | Improved |
| increase latent dimension from 512 to 768 (ablation) | GenSR pretrained model (no CMA-ES) · SRBench noise-free and noisy | R2 and equation complexity | Hurt |
| increase top-k updated dimensions in CMA-ES up to 256 (ablation) | GenSR CMA-ES · SRBench | R2 and search time: R2 saturates at k=256; time grows roughly linearly with k | Improved |
| increase CMA-ES population size (ablation) | GenSR CMA-ES · SRBench | R2: 0.95 -> nearly 0.99, saturating after s=50; runtime tens of seconds -> nearly 300 seconds | Improved |
| increase CMA-ES initial step size (ablation) | GenSR CMA-ES · SRBench | R2: best around t=1.1; complexity grows monotonically | Mixed |
| raise fitness weight omega on complexity (ablation) | GenSR CMA-ES · SRBench | R2 vs equation complexity | Mixed |
| use generative CVAE latent space instead of discriminative E2ESR/SNIP spaces for interpolation | GenSR · Latent linear interpolation | syntactic validity of decoded expressions | Improved |
Closest papers by meaning
- Deep Generative Symbolic Regression (ICLR 2023 · similarity 0.91)
- Ab Initio Nonparametric Variable Selection for Scalable Symbolic Regression with Large $p$ (ICML 2025 · similarity 0.88)
- Symbolic regression via MDLformer-guided search: from minimizing prediction error to minimizing description length (ICLR 2025 · similarity 0.87)
- SNIP: Bridging Mathematical Symbolic and Numeric Realms with Unified Pre-training (ICLR 2024 · similarity 0.87)
- SMILE: Bridging Continuous Optimization and Discrete Symbolic Recovery (NeurIPS 2026 · similarity 0.87)
- FunctionEvolve: Structure-Guided Symbolic Regression with LLMs (NeurIPS 2026 · similarity 0.87)
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.
