Frame Detection on Logistics Frame Detection (real-world)
92AccuracyGPT-4o
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-4oPrompting Strategy=Auto Prompt + RAG (6-shot)2025.12 | 92 | |
| GPT-4o miniPrompting Strategy=Auto Prompt + RAG (6-shot)2025.12 | 88 | |
| GPT-4oPrompting Strategy=Auto Prompt + RAG (3-shot)2025.12 | 88 | |
| LLaMA 3.1 70BPrompting Strategy=Auto Prompt + RAG (6-shot)2025.12 | 87 | |
| Qwen 2.5 72BPrompting Strategy=Auto Prompt + RAG (6-shot)2025.12 | 87 | |
| GPT-4o miniPrompting Strategy=Auto Prompt + RAG (3-shot)2025.12 | 84 | |
| GPT-4oPrompting Strategy=Manual Prompt (6-shot)2025.12 | 84 | |
| GPT-4oPrompting Strategy=Auto Prompt (0-shot)2025.12 | 84 | |
| GPT-4o miniPrompting Strategy=Manual Prompt (6-shot)2025.12 | 83 | |
| LLaMA 3.1 70BPrompting Strategy=Auto Prompt + RAG (3-shot)2025.12 | 81 | |
| Qwen 2.5 72BPrompting Strategy=Auto Prompt + RAG (3-shot)2025.12 | 81 | |
| GPT-4o miniPrompting Strategy=Auto Prompt (0-shot)2025.12 | 80 | |
| LLaMA 3.1 70BPrompting Strategy=Auto Prompt (0-shot)2025.12 | 80 | |
| Qwen 2.5 72BPrompting Strategy=Auto Prompt (0-shot)2025.12 | 80 | |
| Qwen 2.5 72BPrompting Strategy=Manual Prompt (6-shot)2025.12 | 71 | |
| LLaMA 3.1 70BPrompting Strategy=Manual Prompt (6-shot)2025.12 | 70 |