Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code:   DatasetGenerationError
Exception:    ValueError
Message:      Expected object or value
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1816, in _prepare_split_single
                  for key, table in generator:
                                    ^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 613, in wrapped
                  for item in generator(*args, **kwargs):
                              ~~~~~~~~~^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 281, in _generate_tables
                  examples = [ujson_loads(line) for line in batch.splitlines()]
                              ~~~~~~~~~~~^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 20, in ujson_loads
                  return pd.io.json.ujson_loads(*args, **kwargs)
                         ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
              ValueError: Expected object or value
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1683, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1869, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

benchmark
string
seed
int64
state_id
string
task_id
string
step_num
int64
top_m
int64
goal
string
task_action
string
gt
list
image_path
string
goal_candidate_index
int64
band_oracle_candidate_index
int64
full_oracle_candidate_index
int64
method_candidate_index
dict
goal_score
float64
band_oracle_score
float64
full_oracle_score
float64
wm_goal_action_score
float64
wm_gain_vs_goal
float64
headroom
float64
has_headroom
bool
top2_goal_margin
float64
is_low_margin
bool
band_candidate_indices
list
band_candidates
list
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:0
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
0
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.018000000000000002, "y": 0.6900000000000001 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
3
3
0
{ "goal_action": 3, "action_only": 3, "wm_goal_action": 3, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 3, "wm_delta": 3, "wm_delta_zero": 3, "wm_delta_state_shuffle": 3 }
0
0
0
0
0
0
false
0
true
[ 3, 5 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=41, y=747)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 41, "y": 747 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:0
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
0
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.018000000000000002, "y": 0.6900000000000001 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
3
3
0
{ "goal_action": 3, "action_only": 0, "wm_goal_action": 3, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 3, "wm_delta": 7, "wm_delta_zero": 3, "wm_delta_state_shuffle": 3 }
0
0
0
0
0
0
false
0
true
[ 3, 5, 0, 7, 8 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=41, y=747)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 41, "y": 747 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:1
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
1
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.198, "y": 0.356 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
0
0
0
{ "goal_action": 0, "action_only": 2, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 2, "wm_delta_zero": 0, "wm_delta_state_shuffle": 2 }
0
0
0
0
0
0
false
0.000683
true
[ 0, 2 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=41, y=741)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 41, "y": 741 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:1
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
1
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.198, "y": 0.356 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
0
0
0
{ "goal_action": 0, "action_only": 2, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 1, "wm_delta": 6, "wm_delta_zero": 0, "wm_delta_state_shuffle": 1 }
0
0
0
0
0
0
false
0.000683
true
[ 0, 2, 1, 9, 6 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=41, y=741)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 41, "y": 741 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:2
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
2
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.28600000000000003, "y": 0.737 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
8
8
0
{ "goal_action": 8, "action_only": 8, "wm_goal_action": 4, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 4, "wm_delta": 4, "wm_delta_zero": 8, "wm_delta_state_shuffle": 4 }
0
0
0
0
0
0
false
0.045934
true
[ 8, 4 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=41, y=749)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 41, "y": 749 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:2
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
2
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.28600000000000003, "y": 0.737 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
8
8
0
{ "goal_action": 8, "action_only": 8, "wm_goal_action": 4, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 4, "wm_delta": 4, "wm_delta_zero": 8, "wm_delta_state_shuffle": 4 }
0
0
0
0
0
0
false
0.045934
true
[ 8, 4, 6, 9, 1 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=41, y=749)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 41, "y": 749 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
3
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.5720000000000001, "y": 0.612 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
4
4
0
{ "goal_action": 4, "action_only": 4, "wm_goal_action": 4, "wm_goal_action_zero": 4, "wm_goal_action_state_shuffle": 4, "wm_delta": 4, "wm_delta_zero": 4, "wm_delta_state_shuffle": 4 }
0
0
0
0
0
0
false
0.103558
false
[ 4, 0 ]
[ { "candidate_index": 4, "score": 0, "is_first": false, "code": "```python\npyautogui.write(message='xml tools')\n```", "predicted": [ { "type": "write", "params": { "text": "xml tools" }, "metadata": { "bboxes": [] } } ]...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
3
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.5720000000000001, "y": 0.612 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
4
4
0
{ "goal_action": 4, "action_only": 4, "wm_goal_action": 4, "wm_goal_action_zero": 4, "wm_goal_action_state_shuffle": 4, "wm_delta": 4, "wm_delta_zero": 4, "wm_delta_state_shuffle": 4 }
0
0
0
0
0
0
false
0.103558
false
[ 4, 0, 3, 7, 2 ]
[ { "candidate_index": 4, "score": 0, "is_first": false, "code": "```python\npyautogui.write(message='xml tools')\n```", "predicted": [ { "type": "write", "params": { "text": "xml tools" }, "metadata": { "bboxes": [] } } ]...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:4
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
4
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.29, "y": 0.737 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
6
6
0
{ "goal_action": 6, "action_only": 2, "wm_goal_action": 6, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 2, "wm_delta": 2, "wm_delta_zero": 6, "wm_delta_state_shuffle": 6 }
0
0
0
0
0
0
false
0.014823
true
[ 6, 2 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=547, y=886)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 547, "y": 886 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:4
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
4
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.29, "y": 0.737 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
6
6
0
{ "goal_action": 6, "action_only": 2, "wm_goal_action": 6, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 1, "wm_delta": 0, "wm_delta_zero": 6, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.014823
true
[ 6, 2, 0, 1, 4 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=547, y=886)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 547, "y": 886 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:5
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
5
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.332, "y": 0.458 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
3
3
0
{ "goal_action": 3, "action_only": 3, "wm_goal_action": 3, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 4, "wm_delta": 3, "wm_delta_zero": 3, "wm_delta_state_shuffle": 4 }
0
0
0
0
0
0
false
0
true
[ 3, 4 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=1169, y=667)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1169, "y": 667 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:5
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
5
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.332, "y": 0.458 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
3
3
0
{ "goal_action": 3, "action_only": 5, "wm_goal_action": 3, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 4, "wm_delta": 0, "wm_delta_zero": 3, "wm_delta_state_shuffle": 4 }
0
0
0
0
0
0
false
0
true
[ 3, 4, 5, 0, 1 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=1169, y=667)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1169, "y": 667 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:6
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
6
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.28500000000000003, "y": 0.806 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
1
1
0
{ "goal_action": 1, "action_only": 6, "wm_goal_action": 1, "wm_goal_action_zero": 1, "wm_goal_action_state_shuffle": 1, "wm_delta": 6, "wm_delta_zero": 1, "wm_delta_state_shuffle": 6 }
0
0
0
0
0
0
false
0.070915
false
[ 1, 6 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=1049, y=337)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1049, "y": 337 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:6
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
6
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.28500000000000003, "y": 0.806 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
1
1
0
{ "goal_action": 1, "action_only": 6, "wm_goal_action": 1, "wm_goal_action_zero": 1, "wm_goal_action_state_shuffle": 1, "wm_delta": 4, "wm_delta_zero": 1, "wm_delta_state_shuffle": 3 }
0
0
0
0
0
0
false
0.070915
false
[ 1, 6, 8, 4, 3 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=1049, y=337)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1049, "y": 337 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:7
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
7
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.583, "y": 0.612 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
5
5
0
{ "goal_action": 5, "action_only": 5, "wm_goal_action": 5, "wm_goal_action_zero": 5, "wm_goal_action_state_shuffle": 8, "wm_delta": 5, "wm_delta_zero": 5, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.033111
true
[ 5, 8 ]
[ { "candidate_index": 5, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=616, y=590)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 616, "y": 590 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:7
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
7
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.583, "y": 0.612 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
5
5
0
{ "goal_action": 5, "action_only": 5, "wm_goal_action": 5, "wm_goal_action_zero": 5, "wm_goal_action_state_shuffle": 8, "wm_delta": 2, "wm_delta_zero": 5, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.033111
true
[ 5, 8, 2, 4, 7 ]
[ { "candidate_index": 5, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=616, y=590)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 616, "y": 590 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:8
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
8
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.29, "y": 0.803 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
8
8
0
{ "goal_action": 8, "action_only": 5, "wm_goal_action": 8, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 8, "wm_delta": 5, "wm_delta_zero": 8, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.066706
false
[ 8, 5 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=1003, y=336)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1003, "y": 336 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:8
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
8
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.29, "y": 0.803 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
8
8
0
{ "goal_action": 8, "action_only": 4, "wm_goal_action": 8, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 8, "wm_delta": 4, "wm_delta_zero": 8, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.066706
false
[ 8, 5, 4, 9, 7 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=1003, y=336)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1003, "y": 336 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:10
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
10
2
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.318, "y": 0.547 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
0
0
0
{ "goal_action": 0, "action_only": 0, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 0, "wm_delta_zero": 0, "wm_delta_state_shuffle": 2 }
0
0
0
0
0
0
false
0
true
[ 0, 2 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=1007, y=334)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1007, "y": 334 } }, "metadata": { ...
agentnetbench
3
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6:10
01a789bc-03ef-4c40-a7c6-9b1826dc4ed6
10
5
"Please help me install the recommended extensions "XML Tools" and "Rainbow CSV" in VS Code, disable auto-updates for both extensions, and then uninstall the extension "Vue Peek" from the extensions list." <image> What is the next action?
pyautogui.click(x=0.018, y=0.69)
[ { "type": "click", "params": { "position": { "x": 0.318, "y": 0.547 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/22a20375-64bd-4863-a761-c5941e9f518c.png
0
0
0
{ "goal_action": 0, "action_only": 4, "wm_goal_action": 7, "wm_goal_action_zero": 7, "wm_goal_action_state_shuffle": 0, "wm_delta": 4, "wm_delta_zero": 0, "wm_delta_state_shuffle": 2 }
0
0
0
0
0
0
false
0
true
[ 0, 2, 6, 4, 7 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=1007, y=334)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 1007, "y": 334 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:0
01e9b036-ac0c-4e21-ab36-83067cc4cc35
0
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.025, "y": 0.129 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
8
8
0
{ "goal_action": 8, "action_only": 0, "wm_goal_action": 8, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 8, "wm_delta": 0, "wm_delta_zero": 8, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.007723
true
[ 8, 0 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=29, y=136)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 29, "y": 136 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:0
01e9b036-ac0c-4e21-ab36-83067cc4cc35
0
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.025, "y": 0.129 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
8
8
0
{ "goal_action": 8, "action_only": 0, "wm_goal_action": 8, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 8, "wm_delta": 0, "wm_delta_zero": 8, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.007723
true
[ 8, 0, 1, 3, 9 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=29, y=136)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 29, "y": 136 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:1
01e9b036-ac0c-4e21-ab36-83067cc4cc35
1
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.046, "y": 0.069 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
2
2
0
{ "goal_action": 2, "action_only": 2, "wm_goal_action": 2, "wm_goal_action_zero": 2, "wm_goal_action_state_shuffle": 8, "wm_delta": 2, "wm_delta_zero": 2, "wm_delta_state_shuffle": 2 }
0
0
0
0
0
0
false
0.037291
true
[ 2, 8 ]
[ { "candidate_index": 2, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=44, y=133)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 44, "y": 133 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:1
01e9b036-ac0c-4e21-ab36-83067cc4cc35
1
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.046, "y": 0.069 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
2
2
0
{ "goal_action": 2, "action_only": 2, "wm_goal_action": 2, "wm_goal_action_zero": 2, "wm_goal_action_state_shuffle": 8, "wm_delta": 0, "wm_delta_zero": 2, "wm_delta_state_shuffle": 1 }
0
0
0
0
0
0
false
0.037291
true
[ 2, 8, 0, 1, 3 ]
[ { "candidate_index": 2, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=44, y=133)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 44, "y": 133 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:2
01e9b036-ac0c-4e21-ab36-83067cc4cc35
2
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.055, "y": 0.1 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
0
0
0
{ "goal_action": 0, "action_only": 0, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 1, "wm_delta_zero": 0, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.023107
true
[ 0, 1 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=37, y=134)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 37, "y": 134 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:2
01e9b036-ac0c-4e21-ab36-83067cc4cc35
2
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.055, "y": 0.1 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
0
0
0
{ "goal_action": 0, "action_only": 0, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 4, "wm_delta_zero": 0, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.023107
true
[ 0, 1, 8, 9, 4 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=37, y=134)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 37, "y": 134 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:3
01e9b036-ac0c-4e21-ab36-83067cc4cc35
3
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.249, "y": 0.125 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
0
0
0
{ "goal_action": 0, "action_only": 0, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 0, "wm_delta_zero": 0, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0
true
[ 0, 2 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=85, y=76)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 85, "y": 76 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:3
01e9b036-ac0c-4e21-ab36-83067cc4cc35
3
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.249, "y": 0.125 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
0
0
0
{ "goal_action": 0, "action_only": 0, "wm_goal_action": 5, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 5, "wm_delta": 5, "wm_delta_zero": 0, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0
true
[ 0, 2, 4, 8, 5 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=85, y=76)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 85, "y": 76 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:4
01e9b036-ac0c-4e21-ab36-83067cc4cc35
4
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "write", "params": { "text": "python" }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
6
6
0
{ "goal_action": 6, "action_only": 6, "wm_goal_action": 8, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 6, "wm_delta": 8, "wm_delta_zero": 6, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.011892
true
[ 6, 8 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=145, y=104)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 145, "y": 104 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:4
01e9b036-ac0c-4e21-ab36-83067cc4cc35
4
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "write", "params": { "text": "python" }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
6
6
0
{ "goal_action": 6, "action_only": 6, "wm_goal_action": 8, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 5, "wm_delta": 8, "wm_delta_zero": 6, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0.011892
true
[ 6, 8, 9, 1, 5 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=145, y=104)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 145, "y": 104 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:5
01e9b036-ac0c-4e21-ab36-83067cc4cc35
5
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.384, "y": 0.126 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
9
9
0
{ "goal_action": 9, "action_only": 9, "wm_goal_action": 5, "wm_goal_action_zero": 9, "wm_goal_action_state_shuffle": 5, "wm_delta": 5, "wm_delta_zero": 9, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0.006038
true
[ 9, 5 ]
[ { "candidate_index": 9, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=659, y=133)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 659, "y": 133 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:5
01e9b036-ac0c-4e21-ab36-83067cc4cc35
5
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.384, "y": 0.126 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
9
9
0
{ "goal_action": 9, "action_only": 9, "wm_goal_action": 1, "wm_goal_action_zero": 9, "wm_goal_action_state_shuffle": 5, "wm_delta": 5, "wm_delta_zero": 9, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0.006038
true
[ 9, 5, 1, 3, 6 ]
[ { "candidate_index": 9, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=659, y=133)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 659, "y": 133 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:7
01e9b036-ac0c-4e21-ab36-83067cc4cc35
7
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.045, "y": 0.07200000000000001 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
7
7
0
{ "goal_action": 7, "action_only": 7, "wm_goal_action": 7, "wm_goal_action_zero": 7, "wm_goal_action_state_shuffle": 0, "wm_delta": 7, "wm_delta_zero": 7, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.001367
true
[ 7, 0 ]
[ { "candidate_index": 7, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=767, y=134)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 767, "y": 134 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:7
01e9b036-ac0c-4e21-ab36-83067cc4cc35
7
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.045, "y": 0.07200000000000001 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
7
7
0
{ "goal_action": 7, "action_only": 7, "wm_goal_action": 7, "wm_goal_action_zero": 7, "wm_goal_action_state_shuffle": 0, "wm_delta": 7, "wm_delta_zero": 7, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.001367
true
[ 7, 0, 1, 4, 6 ]
[ { "candidate_index": 7, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=767, y=134)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 767, "y": 134 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:8
01e9b036-ac0c-4e21-ab36-83067cc4cc35
8
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.055, "y": 0.358 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
3
3
0
{ "goal_action": 3, "action_only": 9, "wm_goal_action": 3, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 3, "wm_delta": 3, "wm_delta_zero": 3, "wm_delta_state_shuffle": 3 }
0
0
0
0
0
0
false
0.024614
true
[ 3, 9 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=91, y=76)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 91, "y": 76 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:8
01e9b036-ac0c-4e21-ab36-83067cc4cc35
8
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.055, "y": 0.358 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
3
3
0
{ "goal_action": 3, "action_only": 8, "wm_goal_action": 3, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 3, "wm_delta": 3, "wm_delta_zero": 3, "wm_delta_state_shuffle": 3 }
0
0
0
0
0
0
false
0.024614
true
[ 3, 9, 8, 0, 1 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=91, y=76)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 91, "y": 76 } }, "metadata": { ...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:9
01e9b036-ac0c-4e21-ab36-83067cc4cc35
9
2
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.34600000000000003, "y": 0.41100000000000003 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
1
1
0
{ "goal_action": 1, "action_only": 1, "wm_goal_action": 1, "wm_goal_action_zero": 1, "wm_goal_action_state_shuffle": 1, "wm_delta": 7, "wm_delta_zero": 1, "wm_delta_state_shuffle": 1 }
0
0
0
0
0
0
false
0.049624
true
[ 1, 7 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.hotkey(['ctrl', 's'])\n```", "predicted": [ { "type": "hotkey", "params": { "keys": [ "ctrl", "s" ] }, "metadata": { "b...
agentnetbench
3
01e9b036-ac0c-4e21-ab36-83067cc4cc35:9
01e9b036-ac0c-4e21-ab36-83067cc4cc35
9
5
Open VS Code, create a new Python file, write code to implement the "open, read, write" operation of the file, save the file to the "/Documents" folder, run the code, and open the txt document on the desktop to check if text has been written. <image> What is the next action?
pyautogui.click(x=0.025, y=0.129)
[ { "type": "click", "params": { "position": { "x": 0.34600000000000003, "y": 0.41100000000000003 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/34d0363b-60ae-4a1f-9e24-037e80a2bc1d.png
1
1
0
{ "goal_action": 1, "action_only": 1, "wm_goal_action": 1, "wm_goal_action_zero": 1, "wm_goal_action_state_shuffle": 1, "wm_delta": 7, "wm_delta_zero": 1, "wm_delta_state_shuffle": 1 }
0
0
0
0
0
0
false
0.049624
true
[ 1, 7, 0, 2, 3 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.hotkey(['ctrl', 's'])\n```", "predicted": [ { "type": "hotkey", "params": { "keys": [ "ctrl", "s" ] }, "metadata": { "b...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:1
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
1
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.418, "y": 0.131 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
1
1
0
{ "goal_action": 1, "action_only": 1, "wm_goal_action": 1, "wm_goal_action_zero": 1, "wm_goal_action_state_shuffle": 0, "wm_delta": 0, "wm_delta_zero": 1, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.003477
true
[ 1, 0 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=811, y=143)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 811, "y": 143 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:1
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
1
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.418, "y": 0.131 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
1
1
0
{ "goal_action": 1, "action_only": 1, "wm_goal_action": 1, "wm_goal_action_zero": 1, "wm_goal_action_state_shuffle": 0, "wm_delta": 0, "wm_delta_zero": 1, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0.003477
true
[ 1, 0, 2, 5, 6 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=811, y=143)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 811, "y": 143 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:2
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
2
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "press", "params": { "keys": [ "pagedown" ] }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
5
5
0
{ "goal_action": 5, "action_only": 5, "wm_goal_action": 5, "wm_goal_action_zero": 5, "wm_goal_action_state_shuffle": 5, "wm_delta": 5, "wm_delta_zero": 5, "wm_delta_state_shuffle": 2 }
0
0
0
0
0
0
false
0.044197
true
[ 5, 2 ]
[ { "candidate_index": 5, "score": 0, "is_first": false, "code": "```python\ncomputer.tripleClick(x=895, y=402)\n```", "predicted": [], "actions_by_type": {}, "method_scores": { "goal_action": 0.0798924565, "action_only": 0.0795698464, "wm_goal_action": 0.0576276928, ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:2
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
2
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "press", "params": { "keys": [ "pagedown" ] }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
5
5
0
{ "goal_action": 5, "action_only": 5, "wm_goal_action": 5, "wm_goal_action_zero": 5, "wm_goal_action_state_shuffle": 5, "wm_delta": 5, "wm_delta_zero": 5, "wm_delta_state_shuffle": 2 }
0
0
0
0
0
0
false
0.044197
true
[ 5, 2, 1, 7, 0 ]
[ { "candidate_index": 5, "score": 0, "is_first": false, "code": "```python\ncomputer.tripleClick(x=895, y=402)\n```", "predicted": [], "actions_by_type": {}, "method_scores": { "goal_action": 0.0798924565, "action_only": 0.0795698464, "wm_goal_action": 0.0576276928, ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:4
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
4
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.424, "y": 0.128 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
0
0
0
{ "goal_action": 0, "action_only": 0, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 0, "wm_delta_zero": 0, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0
true
[ 0, 2 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=810, y=143)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 810, "y": 143 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:4
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
4
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.424, "y": 0.128 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
0
0
0
{ "goal_action": 0, "action_only": 0, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 1, "wm_delta_zero": 0, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0
true
[ 0, 2, 3, 1, 8 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.click(x=810, y=143)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 810, "y": 143 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:5
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
5
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "moveTo", "params": { "position": { "x": 0.244, "y": 0.653 } }, "metadata": { "bboxes": [] } }, { "type": "dragTo", "params": { "position": { "x": 0.663, "y": 0.758 } }, "metadata": { "bboxes": [] ...
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
0
0
0
{ "goal_action": 0, "action_only": 8, "wm_goal_action": 8, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 8, "wm_delta": 8, "wm_delta_zero": 0, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.000982
true
[ 0, 8 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.press('pagedown')\n```", "predicted": [ { "type": "press", "params": { "keys": [ "pagedown" ] }, "metadata": { "bboxes": [] ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:5
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
5
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "moveTo", "params": { "position": { "x": 0.244, "y": 0.653 } }, "metadata": { "bboxes": [] } }, { "type": "dragTo", "params": { "position": { "x": 0.663, "y": 0.758 } }, "metadata": { "bboxes": [] ...
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
0
0
0
{ "goal_action": 0, "action_only": 8, "wm_goal_action": 8, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 8, "wm_delta": 2, "wm_delta_zero": 0, "wm_delta_state_shuffle": 9 }
0
0
0
0
0
0
false
0.000982
true
[ 0, 8, 4, 9, 2 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.press('pagedown')\n```", "predicted": [ { "type": "press", "params": { "keys": [ "pagedown" ] }, "metadata": { "bboxes": [] ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:6
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
6
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.341, "y": 0.137 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
2
2
0
{ "goal_action": 2, "action_only": 6, "wm_goal_action": 2, "wm_goal_action_zero": 2, "wm_goal_action_state_shuffle": 2, "wm_delta": 2, "wm_delta_zero": 2, "wm_delta_state_shuffle": 6 }
0
0
0
0
0
0
false
0.004793
true
[ 2, 6 ]
[ { "candidate_index": 2, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=659, y=150)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 659, "y": 150 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:6
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
6
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.341, "y": 0.137 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
2
2
0
{ "goal_action": 2, "action_only": 6, "wm_goal_action": 2, "wm_goal_action_zero": 4, "wm_goal_action_state_shuffle": 5, "wm_delta": 4, "wm_delta_zero": 2, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0.004793
true
[ 2, 6, 5, 4, 3 ]
[ { "candidate_index": 2, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=659, y=150)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 659, "y": 150 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:8
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
8
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.343, "y": 0.131 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
8
8
0
{ "goal_action": 8, "action_only": 8, "wm_goal_action": 0, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 8, "wm_delta": 0, "wm_delta_zero": 8, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.04161
true
[ 8, 0 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=659, y=143)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 659, "y": 143 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:8
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
8
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.343, "y": 0.131 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
8
8
0
{ "goal_action": 8, "action_only": 8, "wm_goal_action": 0, "wm_goal_action_zero": 8, "wm_goal_action_state_shuffle": 8, "wm_delta": 0, "wm_delta_zero": 8, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.04161
true
[ 8, 0, 5, 1, 7 ]
[ { "candidate_index": 8, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=659, y=143)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 659, "y": 143 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:9
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
9
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.313, "y": 0.394 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
3
3
0
{ "goal_action": 3, "action_only": 3, "wm_goal_action": 3, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 9, "wm_delta": 3, "wm_delta_zero": 3, "wm_delta_state_shuffle": 9 }
0
0
0
0
0
0
false
0
true
[ 3, 9 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=587, y=428)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 587, "y": 428 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:9
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
9
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.313, "y": 0.394 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
3
3
0
{ "goal_action": 3, "action_only": 3, "wm_goal_action": 4, "wm_goal_action_zero": 3, "wm_goal_action_state_shuffle": 4, "wm_delta": 6, "wm_delta_zero": 3, "wm_delta_state_shuffle": 9 }
0
0
0
0
0
0
false
0
true
[ 3, 9, 4, 6, 1 ]
[ { "candidate_index": 3, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=587, y=428)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 587, "y": 428 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:10
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
10
2
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.51, "y": 0.137 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
6
6
0
{ "goal_action": 6, "action_only": 1, "wm_goal_action": 1, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 6, "wm_delta": 1, "wm_delta_zero": 6, "wm_delta_state_shuffle": 6 }
0
0
0
0
0
0
false
0.032004
true
[ 6, 1 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=653, y=145)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 653, "y": 145 } }, "metadata": { ...
agentnetbench
3
05619e96-d1f5-4f73-9b76-4a8faa6ba48f:10
05619e96-d1f5-4f73-9b76-4a8faa6ba48f
10
5
Add strikethrough to the first two subtitles of the document, change the fourth paragraph of the main text to 16-point green font, then select the content of the fifth paragraph of the main text and insert a bookmark. <image> What is the next action?
computer.tripleClick(x=0.249, y=0.42)
[ { "type": "click", "params": { "position": { "x": 0.51, "y": 0.137 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/867f5a76-92e5-44bc-aced-6cec2f117b8c.png
6
6
0
{ "goal_action": 6, "action_only": 9, "wm_goal_action": 9, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 6, "wm_delta": 1, "wm_delta_zero": 6, "wm_delta_state_shuffle": 6 }
0
0
0
0
0
0
false
0.032004
true
[ 6, 1, 3, 9, 5 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=653, y=145)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 653, "y": 145 } }, "metadata": { ...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:0
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
0
2
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "click", "params": { "position": { "x": 0.495, "y": 0.306 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
6
6
0
{ "goal_action": 6, "action_only": 6, "wm_goal_action": 8, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 8, "wm_delta": 8, "wm_delta_zero": 6, "wm_delta_state_shuffle": 8 }
0
0
0
0
0
0
false
0.012315
true
[ 6, 8 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.moveTo(x=960, y=256)\npyautogui.dragTo(x=960, y=317, button='left')\n```", "predicted": [ { "type": "moveTo", "params": { "position": { "x": 960, "y": 256 ...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:0
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
0
5
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "click", "params": { "position": { "x": 0.495, "y": 0.306 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
6
6
0
{ "goal_action": 6, "action_only": 6, "wm_goal_action": 2, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 9, "wm_delta": 2, "wm_delta_zero": 6, "wm_delta_state_shuffle": 9 }
0
0
0
0
0
0
false
0.012315
true
[ 6, 8, 9, 2, 3 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.moveTo(x=960, y=256)\npyautogui.dragTo(x=960, y=317, button='left')\n```", "predicted": [ { "type": "moveTo", "params": { "position": { "x": 960, "y": 256 ...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:1
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
1
2
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "hotkey", "params": { "keys": [ "ctrl", "f2" ] }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
2
2
0
{ "goal_action": 2, "action_only": 2, "wm_goal_action": 5, "wm_goal_action_zero": 2, "wm_goal_action_state_shuffle": 5, "wm_delta": 2, "wm_delta_zero": 2, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0.021664
true
[ 2, 5 ]
[ { "candidate_index": 2, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=935, y=341)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 935, "y": 341 } }, "metadata": { ...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:1
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
1
5
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "hotkey", "params": { "keys": [ "ctrl", "f2" ] }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
2
2
0
{ "goal_action": 2, "action_only": 2, "wm_goal_action": 5, "wm_goal_action_zero": 2, "wm_goal_action_state_shuffle": 5, "wm_delta": 2, "wm_delta_zero": 2, "wm_delta_state_shuffle": 0 }
0
0
0
0
0
0
false
0.021664
true
[ 2, 5, 3, 0, 7 ]
[ { "candidate_index": 2, "score": 0, "is_first": false, "code": "```python\npyautogui.click(x=935, y=341)\n```", "predicted": [ { "type": "click", "params": { "position": { "x": 935, "y": 341 } }, "metadata": { ...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:2
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
2
2
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "write", "params": { "text": "GEOMEAN" }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
1
1
0
{ "goal_action": 1, "action_only": 1, "wm_goal_action": 2, "wm_goal_action_zero": 2, "wm_goal_action_state_shuffle": 1, "wm_delta": 1, "wm_delta_zero": 1, "wm_delta_state_shuffle": 1 }
0
0
0
0
0
0
false
0.002533
true
[ 1, 2 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.write(message='=GEOMEAN(I2:I5)')\n```", "predicted": [], "actions_by_type": {}, "method_scores": { "goal_action": -0.0182531029, "action_only": 0.0389825702, "wm_goal_action": -0.007996...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:2
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
2
5
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "write", "params": { "text": "GEOMEAN" }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
1
1
0
{ "goal_action": 1, "action_only": 1, "wm_goal_action": 2, "wm_goal_action_zero": 2, "wm_goal_action_state_shuffle": 1, "wm_delta": 1, "wm_delta_zero": 1, "wm_delta_state_shuffle": 1 }
0
0
0
0
0
0
false
0.002533
true
[ 1, 2, 9, 0, 3 ]
[ { "candidate_index": 1, "score": 0, "is_first": false, "code": "```python\npyautogui.write(message='=GEOMEAN(I2:I5)')\n```", "predicted": [], "actions_by_type": {}, "method_scores": { "goal_action": -0.0182531029, "action_only": 0.0389825702, "wm_goal_action": -0.007996...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
3
2
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "doubleClick", "params": { "position": { "x": 0.628, "y": 0.354 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
6
6
0
{ "goal_action": 6, "action_only": 6, "wm_goal_action": 6, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 6, "wm_delta": 6, "wm_delta_zero": 6, "wm_delta_state_shuffle": 6 }
0
0
0
0
0
0
false
0.009111
true
[ 6, 2 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.write(message='=GEOMEAN(')\n```", "predicted": [ { "type": "write", "params": { "text": "=GEOMEAN(" }, "metadata": { "bboxes": [] } } ]...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
3
5
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "doubleClick", "params": { "position": { "x": 0.628, "y": 0.354 } }, "metadata": { "bboxes": [] } } ]
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
6
6
0
{ "goal_action": 6, "action_only": 6, "wm_goal_action": 6, "wm_goal_action_zero": 6, "wm_goal_action_state_shuffle": 6, "wm_delta": 6, "wm_delta_zero": 6, "wm_delta_state_shuffle": 6 }
0
0
0
0
0
0
false
0.009111
true
[ 6, 2, 8, 0, 3 ]
[ { "candidate_index": 6, "score": 0, "is_first": false, "code": "```python\npyautogui.write(message='=GEOMEAN(')\n```", "predicted": [ { "type": "write", "params": { "text": "=GEOMEAN(" }, "metadata": { "bboxes": [] } } ]...
agentnetbench
3
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0:4
24ebf311-a82b-4d8b-8fa6-6e4e1b0b2ff0
4
2
Calculate the geometric mean of Utilization Rate of Clean Energy (%), output it in cell I6, and set it to a number format with three decimal places. Then set the font color to Green1 in libreoffice. Bold cell B2 and set column B to the best fit with default value. <image> What is the next action?
pyautogui.click(x=0.495, y=0.306)
[ { "type": "moveTo", "params": { "position": { "x": 0.494, "y": 0.23600000000000002 } }, "metadata": { "bboxes": [] } }, { "type": "dragTo", "params": { "position": { "x": 0.491, "y": 0.29 } }, "metadata": { "...
/NHNHOME/data/sanghyeok/qwen-cua/data/agentnetbench_built/val_wm_l2_ubuntu/images/9ab991f0-d9e3-4a16-901d-eedff86a6ab4.png
0
0
0
{ "goal_action": 0, "action_only": 5, "wm_goal_action": 0, "wm_goal_action_zero": 0, "wm_goal_action_state_shuffle": 0, "wm_delta": 5, "wm_delta_zero": 0, "wm_delta_state_shuffle": 5 }
0
0
0
0
0
0
false
0.015686
true
[ 0, 5 ]
[ { "candidate_index": 0, "score": 0, "is_first": true, "code": "```python\npyautogui.moveTo(x=1347, y=418)\npyautogui.scroll(-17)\n```", "predicted": [ { "type": "moveTo", "params": { "position": { "x": 1347, "y": 418 } }, ...
End of preview.

No dataset card yet

Downloads last month
27