Hugging Face's logo

Datasets:
superhatk
/
structural-eval-benchmark

Tasks:
Visual Question Answering
Image-to-Text
Languages:
English
Chinese
Size:
n<1K
Tags:
benchmark
visual-reasoning
multimodal
structural-mechanics
image-to-json
visual-question-answering
License:
Dataset card Files
xet
Community
H1W0XXX commited on
Commit
531b10b
·
1 Parent(s): d2a4c38

Add Gemini-3-Pro-Preview results to leaderboard

Browse files
Files changed (2) hide show
  1. README.md +2 -1
  2. README_EN.md +2 -1
README.md CHANGED
@@ -178,7 +178,8 @@ $$
178
 
179
  | 模型 (Model) | 加权准确率 (Weighted Acc) | 平均成功率 (Avg Success Ratio) | 备注 (Notes) |
180
  | :--- | :---: | :---: | :--- |
181
- | **GLM-4.5V** | **16.18%** | 21.25% | Retries: 3 |
 
182
  | **Qwen3-VL-32B-Instruct** | 13.24% | 23.75% | Retries: 3 |
183
 
184
  *注：数据更新于 2026-01-31，基于 v1.0 数据集 (20题).*
 
178
 
179
  | 模型 (Model) | 加权准确率 (Weighted Acc) | 平均成功率 (Avg Success Ratio) | 备注 (Notes) |
180
  | :--- | :---: | :---: | :--- |
181
+ | **Gemini-3-Pro-Preview** | **17.65%** | - | Retries: 3 |
182
+ | **GLM-4.5V** | 16.18% | 21.25% | Retries: 3 |
183
  | **Qwen3-VL-32B-Instruct** | 13.24% | 23.75% | Retries: 3 |
184
 
185
  *注：数据更新于 2026-01-31，基于 v1.0 数据集 (20题).*
README_EN.md CHANGED
@@ -145,7 +145,8 @@ $$
145
 
146
  | Model | Weighted Accuracy | Avg Success Ratio | Notes |
147
  | :--- | :---: | :---: | :--- |
148
- | **GLM-4.5V** | **16.18%** | 21.25% | Retries: 3 |
 
149
  | **Qwen3-VL-32B-Instruct** | 13.24% | 23.75% | Retries: 3 |
150
 
151
  *Note: Data updated on 2026-01-31, based on v1.0 dataset (20 tasks).*
 
145
 
146
  | Model | Weighted Accuracy | Avg Success Ratio | Notes |
147
  | :--- | :---: | :---: | :--- |
148
+ | **Gemini-3-Pro-Preview** | **17.65%** | - | Retries: 3 |
149
+ | **GLM-4.5V** | 16.18% | 21.25% | Retries: 3 |
150
  | **Qwen3-VL-32B-Instruct** | 13.24% | 23.75% | Retries: 3 |
151
 
152
  *Note: Data updated on 2026-01-31, based on v1.0 dataset (20 tasks).*