Social Navigation Perception Prediction on SEAN TOGETHER v2
73CompetenceGemini 2.0 Flash
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Gemini 2.0 FlashCoT=Yes, # Demo. Examples (K)=642025.12 | 73 | 67 | 67 | |
| Gemini 2.0 FlashCoT=Yes, # Demo. Examples (K)=42025.12 | 72 | 64 | 69 | |
| Gemini 2.0 FlashCoT=No, # Demo. Examples (K)=642025.12 | 72 | 70 | 67 | |
| GPT 4.1 miniCoT=No, # Demo. Examples (K)=642025.12 | 71 | 67 | 65 | |
| GPT 4.1 miniCoT=Yes, # Demo. Examples (K)=642025.12 | 70 | 66 | 64 | |
| GPT 4.1 miniCoT=Yes, # Demo. Examples (K)=42025.12 | 69 | 64 | 68 | |
| Gemini 2.0 FlashCoT=No, # Demo. Examples (K)=42025.12 | 67 | 65 | 65 | |
| GPT 4.1 miniCoT=No, # Demo. Examples (K)=42025.12 | 67 | 64 | 65 | |
| RFCoT=/, # Demo. Examples (K)=642025.12 | 66 | 69 | 61 | |
| Llama 3.2 90BCoT=Yes, # Demo. Examples (K)=42025.12 | 61 | 52 | 57 | |
| Llama 3.2 90BCoT=No, # Demo. Examples (K)=42025.12 | 57 | 55 | 51 | |
| RFCoT=/, # Demo. Examples (K)=42025.12 | 53 | 57 | 49 | |
| Llama 3.2 90BCoT=No, # Demo. Examples (K)=642025.12 | 53 | 59 | 45 | |
| Llama 3.2 90BCoT=Yes, # Demo. Examples (K)=642025.12 | 51 | 48 | 51 |