Hugging Face's logo

Datasets:
UniParser
/
RxnBench

Tasks:
Visual Question Answering
Modalities:
Image
Text
Formats:
parquet
Languages:
English
Chinese
Size:
1K - 10K
ArXiv:
Tags:
chemistry
Libraries:
Datasets
pandas
License:
Dataset card Data Studio Files
xet
Community
1
AI4Industry commited on
Commit
35272bb
·
verified ·
1 Parent(s): 07c484f

Add Gemini-3-Pro-preview and GPT-5.1(high)

Browse files
Files changed (1) hide show
  1. README.md +7 -3
README.md CHANGED
@@ -124,9 +124,11 @@ We evaluated several of the latest popular MLLMs, including both closed-source a
124
 
125
  | Moldel | think | Weight | UpdateTime | RxnBench-En | RxnBench-Zh | Mean Score |
126
  | ---- |:----:|:----:|:----:|:----:|:----:|:----:|
127
- | GPT-5(high) | √ | Proprietary | 20250807 | **0.9279** | 0.9246 | **0.9263** |
 
128
  | Gemini-2.5-Pro | √ | Proprietary | 20250617 | 0.9095 | **0.9423** | 0.9259 |
129
- | GPT-5(medium) | √ | Proprietary | 20250807 | 0.9207 | 0.9226 | 0.9217 |
 
130
  | Qwen3-VL-235BA22B-Think | √ | Open | - | 0.9220 | 0.9134 | 0.9177 |
131
  | Qwen3-VL-32B-Think | √ | Open | - | 0.9128 | 0.9161 | 0.9144 |
132
  | GPT-5.1(medium) | √ | Proprietary | 20251113 | 0.9108 | 0.9141 | 0.9125 |
@@ -165,9 +167,11 @@ We also conducted separate evaluations for different task types (in RxnBench-en)
165
 
166
  | Moldel | think | Weight | UpdateTime | Type0 | Type1 | Type2 | Type3 | Type4 | Type5 |
167
  | ---- |:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|
 
168
  | GPT-5(high) | √ | Proprietary | 20250807 | 0.9313 | **0.9444** | **0.9527** | 0.9167 | **0.9661** | 0.8358 |
169
  | Gemini-2.5-Pro | √ | Proprietary | 20250617 | 0.9331 | 0.9246 | 0.9459 | **0.9491** | 0.9322 | 0.6343 |
170
- | GPT-5(medium) | √ | Proprietary | 20250807 | **0.9349** | 0.9325 | 0.9493 | 0.9167 | 0.9492 | 0.7761 |
 
171
  | Qwen3-VL-235BA22B-Think | √ | Open | - | 0.9190 | 0.9405 | 0.9459 | 0.9213 | 0.9322 | **0.8433** |
172
  | Qwen3-VL-32B-Think | √ | Open | - | 0.9296 | 0.9405 | 0.9426 | 0.9259 | 0.9153 | 0.7015 |
173
  | GPT-5.1(medium) | √ | Proprietary | 20251113 | 0.9243 | 0.9365 | 0.9426 | 0.9167 | 0.9492 | 0.7090 |
 
124
 
125
  | Moldel | think | Weight | UpdateTime | RxnBench-En | RxnBench-Zh | Mean Score |
126
  | ---- |:----:|:----:|:----:|:----:|:----:|:----:|
127
+ | Gemini-3-Pro-preview | √ | Proprietary | 20251119 | **0.9318** | 0.9403 | **0.9361** |
128
+ | GPT-5(high) | √ | Proprietary | 20250807 | 0.9279 | 0.9246 | 0.9263 |
129
  | Gemini-2.5-Pro | √ | Proprietary | 20250617 | 0.9095 | **0.9423** | 0.9259 |
130
+ | GPT-5.1(high) | √ | Proprietary | 20251113 | 0.9213 | 0.9220 | 0.9216 |
131
+ | GPT-5(medium) | √ | Proprietary | 20250807 | 0.9207 | 0.9226 | 0.9216 |
132
  | Qwen3-VL-235BA22B-Think | √ | Open | - | 0.9220 | 0.9134 | 0.9177 |
133
  | Qwen3-VL-32B-Think | √ | Open | - | 0.9128 | 0.9161 | 0.9144 |
134
  | GPT-5.1(medium) | √ | Proprietary | 20251113 | 0.9108 | 0.9141 | 0.9125 |
 
167
 
168
  | Moldel | think | Weight | UpdateTime | Type0 | Type1 | Type2 | Type3 | Type4 | Type5 |
169
  | ---- |:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|
170
+ | Gemini-3-Pro-preview | √ | Proprietary | 20251119 | **0.9648** | 0.9246 | **0.9527** | 0.9398 | 0.9322 | 0.7463 |
171
  | GPT-5(high) | √ | Proprietary | 20250807 | 0.9313 | **0.9444** | **0.9527** | 0.9167 | **0.9661** | 0.8358 |
172
  | Gemini-2.5-Pro | √ | Proprietary | 20250617 | 0.9331 | 0.9246 | 0.9459 | **0.9491** | 0.9322 | 0.6343 |
173
+ | GPT-5.1(high) | √ | Proprietary | 20251113 | 0.9243 | 0.9524 | 0.9426 | 0.9167 | 0.9661 | 0.7910 |
174
+ | GPT-5(medium) | √ | Proprietary | 20250807 | 0.9349 | 0.9325 | 0.9493 | 0.9167 | 0.9492 | 0.7761 |
175
  | Qwen3-VL-235BA22B-Think | √ | Open | - | 0.9190 | 0.9405 | 0.9459 | 0.9213 | 0.9322 | **0.8433** |
176
  | Qwen3-VL-32B-Think | √ | Open | - | 0.9296 | 0.9405 | 0.9426 | 0.9259 | 0.9153 | 0.7015 |
177
  | GPT-5.1(medium) | √ | Proprietary | 20251113 | 0.9243 | 0.9365 | 0.9426 | 0.9167 | 0.9492 | 0.7090 |