Symbolic Regression on PO LSR-SYNTH (OOD)
96.6Acc@0.1STRIDE
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| STRIDELLM Backbone=Claude-3-Haiku2026.05 | 96.6 | 60 | 0.0002 | |
| STRIDELLM Backbone=GPT-3.52026.05 | 80 | 80 | 0.0509 | |
| SR-ScientistLLM Backbone=Claude-3-Haiku2026.05 | 75 | 0 | 0.0299 | |
| LLM-SRLLM Backbone=GPT-3.52026.05 | 64.96 | 0 | 743 | |
| LLM-SRLLM Backbone=GPT-3.52026.05 | 60.64 | 60 | 6.94 | |
| STRIDELLM Backbone=Claude-3-Haiku2026.05 | 60.12 | 0 | 0.0106 | |
| STRIDELLM Backbone=GPT-3.52026.05 | 57.56 | 0 | 0.0477 | |
| LLM-SRLLM Backbone=Claude-3-Haiku2026.05 | 48.6 | 0 | 202,000 | |
| SR-ScientistLLM Backbone=GPT-3.52026.05 | 47.56 | 20 | 0.899 | |
| LLM-SRLLM Backbone=Claude-3-Haiku2026.05 | 44.84 | 0 | 212 | |
| SR-ScientistLLM Backbone=GPT-3.52026.05 | 44.48 | 0 | 0.0131 | |
| SR-ScientistLLM Backbone=Claude-3-Haiku2026.05 | 42.28 | 20 | 138 |