Game Interaction on MMG2Skill-Bench Game
83.33Success RateMMG2Skill
Evaluation Results
| Method | Links | |
|---|---|---|
| MMG2SkillBackbone=Claude-Opus-4.6, Prompting Variant=MMG2Skill2026.06 | 83.33 | |
| MMG2SkillBackbone=GPT-5.5, Prompting Variant=MMG2Skill2026.06 | 80 | |
| Raw GuideBackbone=GPT-5.5, Prompting Variant=Raw Guide2026.06 | 76.67 | |
| VanillaBackbone=GPT-5.5, Prompting Variant=Vanilla2026.06 | 73.33 | |
| MMG2SkillBackbone=Claude-Sonnet-4.6, Prompting Variant=MMG2Skill2026.06 | 73.33 | |
| VanillaBackbone=Claude-Opus-4.6, Prompting Variant=Vanilla2026.06 | 66.67 | |
| Raw GuideBackbone=Claude-Opus-4.6, Prompting Variant=Raw Guide2026.06 | 66.67 | |
| MMG2SkillBackbone=Gemini-3.1-Pro-Preview, Prompting Variant=MMG2Skill2026.06 | 66.67 | |
| Raw GuideBackbone=Claude-Sonnet-4.6, Prompting Variant=Raw Guide2026.06 | 60 | |
| VanillaBackbone=Claude-Sonnet-4.6, Prompting Variant=Vanilla2026.06 | 56.67 | |
| MMG2SkillBackbone=Kimi-K2.6, Prompting Variant=MMG2Skill2026.06 | 46.67 | |
| MMG2SkillBackbone=Qwen3.6-Plus, Prompting Variant=MMG2Skill2026.06 | 46.67 | |
| VanillaBackbone=Gemini-3.1-Pro-Preview, Prompting Variant=Vanilla2026.06 | 33.33 | |
| Raw GuideBackbone=Gemini-3.1-Pro-Preview, Prompting Variant=Raw Guide2026.06 | 23.33 | |
| VanillaBackbone=Kimi-K2.6, Prompting Variant=Vanilla2026.06 | 16.67 | |
| Raw GuideBackbone=Kimi-K2.6, Prompting Variant=Raw Guide2026.06 | 16.67 | |
| VanillaBackbone=Qwen3.6-Plus, Prompting Variant=Vanilla2026.06 | 16.67 | |
| Raw GuideBackbone=Qwen3.6-Plus, Prompting Variant=Raw Guide2026.06 | 10 |