Symbolic Regression on LSR-Synth Biology
29.17SAPySR+
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| PySR+Optimization Strategy=SAGE-Fit (Inner-loop)2026.05 | 29.17 | 45.83 | 0.0035 | |
| LLM-SR+Backbone=Qwen2.5-7B, Optimization Strategy=SAGE-Fit (Inner-loop)2026.05 | 20.8 | 41.67 | 0 | |
| LaSR+Backbone=Qwen2.5-7B, Optimization Strategy=SAGE-Fit (Inner-loop)2026.05 | 16.67 | 8.33 | 0.001 | |
| LLM-SRBackbone=Qwen2.5-7B, Optimization Strategy=Post-hoc Refinement2026.05 | 16.67 | 41.67 | 0 | |
| PySROptimization Strategy=Post-hoc Refinement2026.05 | 13.04 | 19.05 | 0.0037 | |
| uDSR+Optimization Strategy=SAGE-Fit (Inner-loop)2026.05 | 4.17 | 23.81 | 0.0237 | |
| PiT-POBackbone=Llama-3.1-8B-Instruct2026.02 | 0.2917 | 0.7083 | 0 | |
| LLM-SRBackbone=Llama-3.1-8B-Instruct2026.02 | 0.253 | 0.5833 | 0 | |
| LaSRBackbone=Llama-3.1-8B-Instruct2026.02 | 0.0416 | 0.1666 | 0.0003 | |
| Direct PromptingBackbone=Llama-3.1-8B-Instruct2026.02 | 0 | 0 | 0.5481 | |
| SGABackbone=Llama-3.1-8B-Instruct2026.02 | 0 | 0 | 0.2416 | |
| uDSROptimization Strategy=Post-hoc Refinement2026.05 | 0 | 20.83 | 0.0776 | |
| LaSRBackbone=Qwen2.5-7B, Optimization Strategy=Post-hoc Refinement2026.05 | 0 | 0 | 0.347 |