Robotic Planning on LEMMA Single-Agent Absence
96Success Rate (SR)SG-CoT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SG-CoTLLM=Gemini-2.5-Flash2026.03 | 96 | 86 | |
| SG-CoTLLM=Qwen3-VL-2B2026.03 | 90 | 80 | |
| SG-CoTLLM=Gemini-2.5-Flash, Scene Graph Edges=false2026.03 | 84 | 49 | |
| InnerMonoLLM=Gemini-2.5-Flash2026.03 | 81 | 48 | |
| SG-CoTLLM=Gemini-2.5-Flash, Iterations=12026.03 | 71 | 45 | |
| SG-CoTLLM=Qwen3-VL-2B, Scene Graph Edges=false2026.03 | 69 | 41 | |
| InnerMonoLLM=Qwen3-VL-2B2026.03 | 67 | 52 | |
| SG-CoTLLM=Qwen3-VL-2B, Iterations=12026.03 | 60 | 42 | |
| ProgPromptLLM=Gemini-2.5-Flash2026.03 | 48 | 31 | |
| SG-CoTLLM=Gemini-2.5-Flash, Iterations=02026.03 | 44 | 36 | |
| CLARALLM=Gemini-2.5-Flash2026.03 | 37 | 33 | |
| CLARALLM=Qwen3-VL-2B2026.03 | 35 | 35 | |
| ProgPromptLLM=Qwen3-VL-2B2026.03 | 33 | 28 | |
| SG-CoTLLM=Qwen3-VL-2B, Iterations=02026.03 | 32 | 26 |