Code Generation on MBPP+ Intermediate persona
90Win-rateGemini-3-Pro vs. Gemma-3-4B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemini-3-Pro vs. Gemma-3-4BPrompt Type=Personalized2026.04 | 90 | 1 | |
| GPT-5.1 vs. GPT-4oPrompt Type=Personalized2026.04 | 77 | 2 | |
| Qwen3-32B vs. Qwen3-14BPrompt Type=Personalized2026.04 | 67 | 4 | |
| Qwen3-32B vs. Qwen3-14BPrompt Type=Original2026.04 | 62 | 3 | |
| Qwen3-32B vs. Qwen3-14BPrompt Type=Control2026.04 | 59 | 4 | |
| GPT-5.1 vs. GPT-OSS-20BPrompt Type=Personalized2026.04 | 55 | 2 | |
| Gemini-3-Pro vs. Gemma-3-4BPrompt Type=Original2026.04 | 53 | 1 | |
| Gemini-3-Pro vs. Gemma-3-4BPrompt Type=Control2026.04 | 45 | 2 | |
| GPT-5.1 vs. GPT-4oPrompt Type=Control2026.04 | 19 | 3 | |
| GPT-5.1 vs. GPT-4oPrompt Type=Original2026.04 | 16 | 2 | |
| GPT-5.1 vs. GPT-OSS-20BPrompt Type=Control2026.04 | 2 | 0 | |
| GPT-5.1 vs. GPT-OSS-20BPrompt Type=Original2026.04 | 1 | 0 |