Web UI Question Answering on WebSRC
0.886AccuracyDAVE
Evaluation Results
| Method | Links | |
|---|---|---|
| DAVELLM Backbone=Qwen-2.5-7B-Instruct2025.12 | 0.886 | |
| SigLIP 2LLM Backbone=Qwen-2.5-7B-Instruct2025.12 | 0.883 | |
| AIMv2LLM Backbone=Qwen-2.5-7B-Instruct2025.12 | 0.861 | |
| DophineLLM Backbone=Qwen-2.5-7B-Instruct2025.12 | 0.856 | |
| DAVELLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.826 | |
| DophineLLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.76 | |
| SigLIP 2LLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.678 | |
| MAE-WebLLM Backbone=Qwen-2.5-7B-Instruct2025.12 | 0.607 | |
| AIMv2LLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.494 | |
| MAE-WebLLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.468 | |
| Pix2structLLM Backbone=Qwen-2.5-7B-Instruct2025.12 | 0.303 | |
| Pix2StructLLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.265 | |
| DinoV2LLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.176 | |
| DiTLLM Backbone=Llama-3.2-3B-Instruct2025.12 | 0.149 |