Character Probing on ARCANE In-Scenario The Underdogs, East Lynne (low-popularity unvalidated)
64.7APFDeepSeek-V4 Pro Arc
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DeepSeek-V4 Pro ArcFamily=DeepSeek-V4, Size=Pro, Mode=Arc2026.06 | 64.7 | 63.6 | 57.1 | 60.5 | |
| DeepSeek-V4 Pro LifeChoiceFamily=DeepSeek-V4, Size=Pro, Mode=LifeChoice2026.06 | 64.4 | 63.8 | 57.8 | 58.7 | |
| DeepSeek-V4 Pro RAGFamily=DeepSeek-V4, Size=Pro, Mode=RAG2026.06 | 64.1 | 63.1 | 57.6 | 58.4 | |
| DeepSeek-V4 Flash ArcFamily=DeepSeek-V4, Size=Flash, Mode=Arc2026.06 | 63.4 | 63 | 55.3 | 60.9 | |
| DeepSeek-V4 Flash RAGFamily=DeepSeek-V4, Size=Flash, Mode=RAG2026.06 | 62.7 | 61.9 | 55.6 | 59.7 | |
| ARCANE 32B RAGFamily=ARCANE, Size=32B, Mode=RAG2026.06 | 61.7 | 60 | 53.9 | 53.8 | |
| DeepSeek-V4 Flash LifeChoiceFamily=DeepSeek-V4, Size=Flash, Mode=LifeChoice2026.06 | 61 | 60.5 | 53.3 | 52.8 | |
| DeepSeek-V4 Flash SummaryFamily=DeepSeek-V4, Size=Flash, Mode=Summary2026.06 | 60.6 | 60.2 | 52.8 | 56.9 | |
| DeepSeek-V4 Pro SummaryFamily=DeepSeek-V4, Size=Pro, Mode=Summary2026.06 | 60 | 60.2 | 52.3 | 55.7 | |
| ARCANE 32B ArcFamily=ARCANE, Size=32B, Mode=Arc2026.06 | 60 | 58.6 | 52.1 | 56.5 | |
| ARCANE 8B RAGFamily=ARCANE, Size=8B, Mode=RAG2026.06 | 59.1 | 57.6 | 51.9 | 51.2 | |
| ARCANE 8B ArcFamily=ARCANE, Size=8B, Mode=Arc2026.06 | 58.5 | 57.2 | — | 51.6 | |
| ARCANE 32B LifeChoiceFamily=ARCANE, Size=32B, Mode=LifeChoice2026.06 | 57.8 | 55.9 | 49.8 | 53.2 | |
| DeepSeek-V4 Flash VanillaFamily=DeepSeek-V4, Size=Flash, Mode=Vanilla2026.06 | 56.9 | 57.7 | 47.9 | 51.3 | |
| DeepSeek-V4 Flash TimeCharaFamily=DeepSeek-V4, Size=Flash, Mode=TimeChara2026.06 | 56.6 | 57.3 | 48.6 | 51.8 | |
| ARCANE 32B SummaryFamily=ARCANE, Size=32B, Mode=Summary2026.06 | 55.3 | 54.3 | 46.3 | 52.8 | |
| Qwen3 32B RAGFamily=Qwen3, Size=32B, Mode=RAG2026.06 | 55.2 | 53.2 | 45.4 | 41.3 | |
| Qwen3 32B ArcFamily=Qwen3, Size=32B, Mode=Arc2026.06 | 55.1 | 53.5 | 44 | 50.7 | |
| DeepSeek-V4 Pro TimeCharaFamily=DeepSeek-V4, Size=Pro, Mode=TimeChara2026.06 | 54.9 | 55.6 | 46.8 | 50.4 | |
| Qwen3 32B LifeChoiceFamily=Qwen3, Size=32B, Mode=LifeChoice2026.06 | 54.8 | 54.1 | 46 | 45.9 | |
| DeepSeek-V4 Pro VanillaFamily=DeepSeek-V4, Size=Pro, Mode=Vanilla2026.06 | 54.3 | 55.2 | 45.9 | 51.1 | |
| ARCANE 8B LifeChoiceFamily=ARCANE, Size=8B, Mode=LifeChoice2026.06 | 53.7 | 52.6 | 45 | 49 | |
| ARCANE 8B SummaryFamily=ARCANE, Size=8B, Mode=Summary2026.06 | 52.8 | 51.4 | 44.5 | 46.1 | |
| Qwen3 32B SummaryFamily=Qwen3, Size=32B, Mode=Summary2026.06 | 50.5 | 50 | 40.3 | 42.7 | |
| Qwen3 8B RAGFamily=Qwen3, Size=8B, Mode=RAG2026.06 | 49.2 | 48 | 39.8 | 37.7 | |
| ARCANE 32B VanillaFamily=ARCANE, Size=32B, Mode=Vanilla2026.06 | 48.9 | 49.3 | 39.9 | 41.6 | |
| ARCANE 32B TimeCharaFamily=ARCANE, Size=32B, Mode=TimeChara2026.06 | 47.4 | 47.9 | 38.6 | 41.6 | |
| Qwen3 8B LifeChoiceFamily=Qwen3, Size=8B, Mode=LifeChoice2026.06 | 47.2 | 45.6 | 37 | 42 | |
| Qwen3 8B ArcFamily=Qwen3, Size=8B, Mode=Arc2026.06 | 46.7 | 46.6 | 36.1 | 37.5 | |
| ARCANE 8B TimeCharaFamily=ARCANE, Size=8B, Mode=TimeChara2026.06 | 44.4 | 45.1 | 36 | 38.6 | |
| ARCANE 8B VanillaFamily=ARCANE, Size=8B, Mode=Vanilla2026.06 | 44.2 | 43.6 | 34.9 | 38.3 | |
| Qwen3 8B SummaryFamily=Qwen3, Size=8B, Mode=Summary2026.06 | 43.3 | 42.3 | 33 | 33.5 | |
| Qwen3 32B TimeCharaFamily=Qwen3, Size=32B, Mode=TimeChara2026.06 | 42.6 | 42.9 | 32.7 | 34.8 | |
| Qwen3 32B VanillaFamily=Qwen3, Size=32B, Mode=Vanilla2026.06 | 40.1 | 40.4 | 30 | 33.2 | |
| Qwen3 8B TimeCharaFamily=Qwen3, Size=8B, Mode=TimeChara2026.06 | 38.7 | 40.3 | 29.5 | 31 | |
| Qwen3 8B VanillaFamily=Qwen3, Size=8B, Mode=Vanilla2026.06 | 35.5 | 36.7 | 26.1 | 27.8 |