Agent Behavior Evaluation on Social Simulation Family context 1.0
4.905NaturalnessASVO
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ASVOLLM=GPT-52026.03 | 4.905 | 4.024 | |
| ASVOLLM=Gemini-2.52026.03 | 4.863 | 4.077 | |
| ASVOLLM=Deepseek-v32026.03 | 4.75 | 3.964 | |
| D2ALLM=Qwen3-235b2026.03 | 4.49 | 3.781 | |
| D2ALLM=GPT-52026.03 | 4.479 | 3.75 | |
| ASVOLLM=Qwen3-235b2026.03 | 4.381 | 3.82 | |
| LLMobLLM=GPT-52026.03 | 4.202 | 3.863 | |
| D2ALLM=Gemini-2.52026.03 | 4.031 | 3.615 | |
| D2ALLM=Deepseek-v32026.03 | 4.01 | 3.167 | |
| ReActLLM=Qwen3-235b2026.03 | 4.006 | 3.536 | |
| LLMobLLM=Qwen3-235b2026.03 | 4 | 3.738 | |
| LLMobLLM=Gemini-2.52026.03 | 3.845 | 3.429 | |
| ReActLLM=Gemini-2.52026.03 | 3.726 | 3.488 | |
| BabyAGILLM=Qwen3-235b2026.03 | 3.685 | 3.315 | |
| LLMobLLM=Deepseek-v32026.03 | 3.589 | 3.339 | |
| BabyAGILLM=Gemini-2.52026.03 | 3.423 | 1.433 | |
| ReActLLM=GPT-52026.03 | 3.345 | 3.202 | |
| BabyAGILLM=GPT-52026.03 | 3.345 | 3.202 | |
| ReActLLM=Deepseek-v32026.03 | 3.286 | 3.262 | |
| BabyAGILLM=Deepseek-v32026.03 | 2.917 | 2.69 |