Character Relationship Network Updating on BookWorld SNP Existing scenario
88.3Win RateDeepSeek-V3.2-Think
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DeepSeek-V3.2-ThinkBaseline=vs. BookWorld (SNP)2026.04 | 88.3 | 3.97 | |
| Gemini-2.5-ProBaseline=vs. BookWorld (SNP)2026.04 | 80 | 3.96 | |
| Qwen3-32BBaseline=vs. BookWorld (SNP)2026.04 | 71.7 | 3.14 | |
| DeepSeek-V3.2Baseline=vs. BookWorld (SNP)2026.04 | 70 | 3.52 | |
| Llama3.3-70BBaseline=vs. BookWorld (SNP)2026.04 | 41.7 | 2.05 |