SQL Code Generation Agent on InterCodeSQL
61.3Average RewardIPR
Evaluation Results
| Method | Links | |
|---|---|---|
| IPRRefinement Category=Process Refinement, Backbone=Llama-2-7B2025.12 | 61.3 | |
| Step-PPORefinement Category=Process Refinement, Backbone=Llama-2-7B2025.12 | 60.2 | |
| MACLARefinement Category=Process Refinement, Backbone=Llama-2-7B2025.12 | 59.3 | |
| ETORefinement Category=Outcome Refinement, Backbone=Llama-2-7B2025.12 | 57.2 | |
| RFT-CRRefinement Category=Outcome Refinement, Backbone=Llama-2-7B2025.12 | 56.3 | |
| SFTRefinement Category=Outcome Refinement, Backbone=Llama-2-7B2025.12 | 54.9 | |
| RFT-PPORefinement Category=Outcome Refinement, Backbone=Llama-2-7B2025.12 | 52.4 | |
| GPT-4Refinement Category=Prompt-based, Backbone=GPT-42025.12 | 38.5 | |
| GPT-3.5-TurboRefinement Category=Prompt-based, Backbone=GPT-3.5-Turbo2025.12 | 37.8 | |
| Llama-2-7BRefinement Category=Prompt-based, Backbone=Llama-2-7B2025.12 | 4 |