Resistance simulation on RPC (test)
70.38PrecisionResistClient
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| ResistClient2026.04 | 70.38 | 78.95 | 74.42 | 38.03 | 1.63 | 1.58 | 2.61 | |
| Qwen3-8B-SFTmode=Supervised Fine-Tuning (SFT)2026.04 | 63.54 | 73.9 | 68.33 | 39.43 | 1.46 | 1.41 | 2.39 | |
| GPT-5.1mode=Prompt-only2026.04 | 59.31 | 62.88 | 61.04 | 35.94 | 1.42 | 1.35 | 2.52 | |
| DeepSeek-V3.2mode=Prompt-only2026.04 | 52.87 | 57.56 | 55.12 | 36.91 | 1.29 | 1.21 | 2.4 | |
| GLM-4.6mode=Prompt-only2026.04 | 49.1 | 54.72 | 51.76 | 37.78 | 1.27 | 1.23 | 2.34 | |
| Kimi-K2-thinkingmode=Prompt-only2026.04 | 47.08 | 51.19 | 49.05 | 36.86 | 1.24 | 1.19 | 2.18 | |
| Qwen3-8Bmode=Prompt-only2026.04 | 36.52 | 48.54 | 41.68 | 45.06 | 1.1 | 1.04 | 1.88 | |
| DeepSeek-R1-8Bmode=Prompt-only2026.04 | 34.67 | 46.26 | 39.64 | 45.23 | 0.98 | 1.08 | 1.72 |