Dataset Preview
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: CastError
Message: Couldn't cast
episode: string
model: string
checkpoint: string
sampling: struct<temperature: double, top_p: double, top_k: int64, thinking: bool, max_output_tokens: int64, i (... 18 chars omitted)
child 0, temperature: double
child 1, top_p: double
child 2, top_k: int64
child 3, thinking: bool
child 4, max_output_tokens: int64
child 5, image_width: int64
budget: struct<actions: int64, model_calls: int64, tool_calls: int64>
child 0, actions: int64
child 1, model_calls: int64
child 2, tool_calls: int64
format: string
world: string
round: int64
family: string
code: string
variant: int64
spawn: string
n_actions: int64
model_calls: int64
done: bool
reported: int64
success: bool
success_clean: bool
tools: null
messages: null
score: null
orig_score: null
neg_refs: list<item: string>
child 0, item: string
source: string
area: struct<minX: double, maxX: double, minZ: double, maxZ: double>
child 0, minX: double
child 1, maxX: double
child 2, minZ: double
child 3, maxZ: double
provider: string
film_stale: bool
neg_inspect: bool
relabel_of: string
masked_turns: int64
source_set: string
neg_region: list<item: double>
child 0, item: double
thought_status: string
neg_action: string
masked_invalid: int64
method_version: string
uid: string
run: string
task: string
film_inspects: int64
neg_step: int64
cut: bool
neg_line: string
neg_object: string
hint_mismatch: bool
profile: string
llm: struct<provider: string, model: string, reasoning: string, max_out: int64, temperature: double, pool (... 51 chars omitted)
child 0, provider: string
child 1, model: string
child 2, reasoning: string
child 3, max_out: int64
child 4, temperature: double
child 5, pool: string
child 6, keep_images: int64
child 7, compact_block: int64
gen_source: string
require_full_budget: bool
prefix_text: string
id: string
arm: string
neg_flag: struct<status: string, category: string, evidence: list<item: string>, description: string>
child 0, status: string
child 1, category: string
child 2, evidence: list<item: string>
child 0, item: string
child 3, description: string
to
{'id': Value('string'), 'task': Value('string'), 'arm': Value('string'), 'world': Value('string'), 'family': Value('string'), 'code': Value('string'), 'budget': Value('int64'), 'profile': Value('string'), 'format': Value('string'), 'provider': Value('string'), 'model': Value('string'), 'hint_mismatch': Value('bool'), 'source': Value('string'), 'thought_status': Value('string'), 'run': Value('string'), 'uid': Value('string'), 'spawn': Json(decode=True), 'area': {'minX': Value('float64'), 'maxX': Value('float64'), 'minZ': Value('float64'), 'maxZ': Value('float64')}, 'require_full_budget': Value('bool'), 'method_version': Value('string'), 'cut': Value('bool'), 'n_actions': Value('int64'), 'masked_turns': Value('int64'), 'tools': {'move': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': {'distance_m': {'type': Value('string'), 'minimum': Value('float64'), 'maximum': Value('int64'), 'description': Value('string')}, 'direction': {'type': Value('string'), 'enum': List(Value('string'))}}, 'required': List(Value('string')), 'additionalProperties': Value('bool')}}, 'turn': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': {'degrees': {'type': Value('string'), 'minimum': Value('int64'), 'maximum': Value('int64'), 'description': Value('string')}}, 'required': List(Value('string')), 'additionalProperties': Value('bool')}}, 'look': {'name': Value('string'), 'descript
...
g'), 'properties': {'text': {'type': Value('string')}}, 'required': List(Value('string')), 'additionalProperties': Value('bool')}}, 'observe': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': Json(decode=True), 'additionalProperties': Value('bool')}}, 'read_example': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': {'code': {'type': Value('string'), 'enum': List(Value('string'))}}, 'additionalProperties': Value('bool')}}}, 'messages': List(Json(decode=True)), 'score': Json(decode=True), 'source_set': Value('string'), 'orig_score': Json(decode=True), 'neg_inspect': Value('bool'), 'relabel_of': Value('string'), 'gen_source': Value('string'), 'neg_flag': {'status': Value('string'), 'category': Value('string'), 'evidence': List(Value('string')), 'description': Value('string')}, 'neg_step': Value('int64'), 'neg_action': Value('string'), 'neg_refs': List(Value('string')), 'neg_region': List(Value('float64')), 'neg_object': Value('string'), 'prefix_text': Value('string'), 'neg_line': Value('string'), 'variant': Value('int64'), 'llm': {'provider': Value('string'), 'model': Value('string'), 'reasoning': Value('string'), 'max_out': Value('int64'), 'temperature': Value('float64'), 'pool': Value('string'), 'keep_images': Value('int64'), 'compact_block': Value('int64')}, 'film_stale': Value('bool'), 'film_inspects': Value('int64'), 'masked_invalid': Value('int64')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1835, in _prepare_split_single
for key, table in generator:
^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 621, in wrapped
for item in generator(*args, **kwargs):
~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
episode: string
model: string
checkpoint: string
sampling: struct<temperature: double, top_p: double, top_k: int64, thinking: bool, max_output_tokens: int64, i (... 18 chars omitted)
child 0, temperature: double
child 1, top_p: double
child 2, top_k: int64
child 3, thinking: bool
child 4, max_output_tokens: int64
child 5, image_width: int64
budget: struct<actions: int64, model_calls: int64, tool_calls: int64>
child 0, actions: int64
child 1, model_calls: int64
child 2, tool_calls: int64
format: string
world: string
round: int64
family: string
code: string
variant: int64
spawn: string
n_actions: int64
model_calls: int64
done: bool
reported: int64
success: bool
success_clean: bool
tools: null
messages: null
score: null
orig_score: null
neg_refs: list<item: string>
child 0, item: string
source: string
area: struct<minX: double, maxX: double, minZ: double, maxZ: double>
child 0, minX: double
child 1, maxX: double
child 2, minZ: double
child 3, maxZ: double
provider: string
film_stale: bool
neg_inspect: bool
relabel_of: string
masked_turns: int64
source_set: string
neg_region: list<item: double>
child 0, item: double
thought_status: string
neg_action: string
masked_invalid: int64
method_version: string
uid: string
run: string
task: string
film_inspects: int64
neg_step: int64
cut: bool
neg_line: string
neg_object: string
hint_mismatch: bool
profile: string
llm: struct<provider: string, model: string, reasoning: string, max_out: int64, temperature: double, pool (... 51 chars omitted)
child 0, provider: string
child 1, model: string
child 2, reasoning: string
child 3, max_out: int64
child 4, temperature: double
child 5, pool: string
child 6, keep_images: int64
child 7, compact_block: int64
gen_source: string
require_full_budget: bool
prefix_text: string
id: string
arm: string
neg_flag: struct<status: string, category: string, evidence: list<item: string>, description: string>
child 0, status: string
child 1, category: string
child 2, evidence: list<item: string>
child 0, item: string
child 3, description: string
to
{'id': Value('string'), 'task': Value('string'), 'arm': Value('string'), 'world': Value('string'), 'family': Value('string'), 'code': Value('string'), 'budget': Value('int64'), 'profile': Value('string'), 'format': Value('string'), 'provider': Value('string'), 'model': Value('string'), 'hint_mismatch': Value('bool'), 'source': Value('string'), 'thought_status': Value('string'), 'run': Value('string'), 'uid': Value('string'), 'spawn': Json(decode=True), 'area': {'minX': Value('float64'), 'maxX': Value('float64'), 'minZ': Value('float64'), 'maxZ': Value('float64')}, 'require_full_budget': Value('bool'), 'method_version': Value('string'), 'cut': Value('bool'), 'n_actions': Value('int64'), 'masked_turns': Value('int64'), 'tools': {'move': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': {'distance_m': {'type': Value('string'), 'minimum': Value('float64'), 'maximum': Value('int64'), 'description': Value('string')}, 'direction': {'type': Value('string'), 'enum': List(Value('string'))}}, 'required': List(Value('string')), 'additionalProperties': Value('bool')}}, 'turn': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': {'degrees': {'type': Value('string'), 'minimum': Value('int64'), 'maximum': Value('int64'), 'description': Value('string')}}, 'required': List(Value('string')), 'additionalProperties': Value('bool')}}, 'look': {'name': Value('string'), 'descript
...
g'), 'properties': {'text': {'type': Value('string')}}, 'required': List(Value('string')), 'additionalProperties': Value('bool')}}, 'observe': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': Json(decode=True), 'additionalProperties': Value('bool')}}, 'read_example': {'name': Value('string'), 'description': Value('string'), 'parameters': {'type': Value('string'), 'properties': {'code': {'type': Value('string'), 'enum': List(Value('string'))}}, 'additionalProperties': Value('bool')}}}, 'messages': List(Json(decode=True)), 'score': Json(decode=True), 'source_set': Value('string'), 'orig_score': Json(decode=True), 'neg_inspect': Value('bool'), 'relabel_of': Value('string'), 'gen_source': Value('string'), 'neg_flag': {'status': Value('string'), 'category': Value('string'), 'evidence': List(Value('string')), 'description': Value('string')}, 'neg_step': Value('int64'), 'neg_action': Value('string'), 'neg_refs': List(Value('string')), 'neg_region': List(Value('float64')), 'neg_object': Value('string'), 'prefix_text': Value('string'), 'neg_line': Value('string'), 'variant': Value('int64'), 'llm': {'provider': Value('string'), 'model': Value('string'), 'reasoning': Value('string'), 'max_out': Value('int64'), 'temperature': Value('float64'), 'pool': Value('string'), 'keep_images': Value('int64'), 'compact_block': Value('int64')}, 'film_stale': Value('bool'), 'film_inspects': Value('int64'), 'masked_invalid': Value('int64')}
because column names don't match
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 1379, 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 956, 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 1702, 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 1887, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
id string | task string | arm string | world string | family string | code string | budget int64 | profile string | format string | provider string | model string | hint_mismatch bool | source string | thought_status string | run string | uid string | spawn unknown | area dict | require_full_budget bool | method_version string | cut bool | n_actions int64 | masked_turns int64 | tools dict | messages list | score unknown | source_set string | orig_score unknown | neg_inspect bool | relabel_of string | gen_source string | neg_flag dict | neg_step int64 | neg_action string | neg_refs list | neg_region list | neg_object string | prefix_text string | neg_line string | variant int64 | llm dict | film_stale bool | film_inspects int64 | masked_invalid int64 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
pl-market-002-r4-nosolid-4:solo | pl-market-002-r4-nosolid-4 | solo | pl-market-002 | nosolid | C1 | 40 | env | mcp-v1 | script | planner-pxqc-v1 | false | script | ok | script3b5 | script3b5/pl-market-002-r4-nosolid-4:solo | {
"pos": [
-3.3,
16.95
],
"yaw_rad": 0.8270000000000001,
"how": "mid"
} | {
"minX": -7.986,
"maxX": 16.014,
"minZ": -2.182,
"maxZ": 21.818
} | false | method_v2 | false | 37 | 0 | {
"move": {
"name": "move",
"description": "Walk in a straight line. Stops early if something blocks the way.",
"parameters": {
"type": "object",
"properties": {
"distance_m": {
"type": "number",
"minimum": 0.3,
"maximum": 4,
"description": "metr... | [
{
"role": "system",
"text": "You are a game QA tester controlling a first-person agent in a 3D world through tools. Your job is to inspect the world for BUGS - anything broken, impossible, or inconsistent with how a normal game world should look and behave - and to report them precisely.\n\nStandard game-bu... | {
"task": "pl-market-002-r4-nosolid-4",
"arm": "solo",
"code": "C1",
"family": "nosolid",
"profile": "env",
"capturable": true,
"reported": 1,
"flags_total": 1,
"retracted": 0,
"success_lenient": true,
"success": true,
"success_clean": true,
"extra_reports": 0,
"first_flag_action": 36,
"n_... | script:script3_success_clean.jsonl | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
pl-street-001-r4-noshadow-0:solo:neg1 | pl-street-001-r4-noshadow-0 | solo | pl-street-001 | noshadow | V3 | 40 | wilds | mcp-v1 | gemini | gemini-3.8-flash | false | neg | null | scale5 | scale5/pl-street-001-r4-noshadow-0:solo:neg1 | {
"pos": [
1.9,
1.85
],
"yaw_rad": -2.452,
"how": "mid"
} | {
"minX": -17.1015,
"maxX": 6.8985,
"minZ": -5.9725,
"maxZ": 18.0275
} | false | method_v2 | false | 37 | 33 | {
"move": {
"name": "move",
"description": "Walk in a straight line. Stops early if something blocks the way.",
"parameters": {
"type": "object",
"properties": {
"distance_m": {
"type": "number",
"minimum": 0.3,
"maximum": 4,
"description": "metr... | [
{
"role": "system",
"text": "You are a game QA tester controlling a first-person agent in a 3D world through tools. Your job is to inspect the world for BUGS - anything broken, impossible, or inconsistent with how a normal game world should look and behave - and to report them precisely.\n\nStandard game-bu... | {
"task": "pl-street-001-r4-noshadow-0",
"arm": "solo",
"code": "V3",
"family": "noshadow",
"profile": "wilds",
"capturable": false,
"reported": 0,
"flags_total": 0,
"retracted": 0,
"success_lenient": false,
"success": false,
"success_clean": false,
"extra_reports": 0,
"first_flag_action": n... | neg1 | {
"task": "pl-street-001-r4-noshadow-0",
"arm": "solo",
"code": "V3",
"family": "noshadow",
"profile": "wilds",
"capturable": false,
"reported": 1,
"flags_total": 1,
"retracted": 0,
"success_lenient": false,
"success": false,
"success_clean": false,
"extra_reports": 1,
"first_flag_action": 3... | true | scale5/pl-street-001-r4-noshadow-0:solo | teacher | {
"status": "confirmed",
"category": "visual",
"evidence": [
"a37",
"a0",
"a32"
],
"description": "The outdoor lamppost on the sidewalk in front of the shop has a detached and displaced shadow: its shadow appears separated on the road/curb to the left, with no shadow attached to the base of the la... | 33 | done | [
"a32"
] | null | null | none | gemini | null | null | null | null | null |
pl-house-002-r4-upside-0:solo | pl-house-002-r4-upside-0 | solo | pl-house-002 | upside | S2 | 40 | cottage | mcp-v1 | script | planner-pxqc-v1 | false | script | ok | script3b3 | script3b3/pl-house-002-r4-upside-0:solo | {
"pos": [
4.8,
-0.8
],
"yaw_rad": 0.357,
"how": "mid"
} | {
"minX": -7.268,
"maxX": 8.35,
"minZ": -2.9,
"maxZ": 17.772
} | false | method_v2 | false | 20 | 0 | {
"move": {
"name": "move",
"description": "Walk in a straight line. Stops early if something blocks the way.",
"parameters": {
"type": "object",
"properties": {
"distance_m": {
"type": "number",
"minimum": 0.3,
"maximum": 4,
"description": "metr... | [
{
"role": "system",
"text": "You are a game QA tester controlling a first-person agent in a 3D world through tools. Your job is to inspect the world for BUGS - anything broken, impossible, or inconsistent with how a normal game world should look and behave - and to report them precisely.\n\nStandard game-bu... | {
"task": "pl-house-002-r4-upside-0",
"arm": "solo",
"code": "S2",
"family": "upside",
"profile": "cottage",
"capturable": true,
"reported": 1,
"flags_total": 1,
"retracted": 0,
"success_lenient": true,
"success": true,
"success_clean": true,
"extra_reports": 0,
"first_flag_action": 10,
"n... | script:script3_success_clean.jsonl | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
pl-school-008-r5-turn-2:solo | pl-school-008-r5-turn-2 | solo | pl-school-008 | turn | T2 | 40 | env | mcp-v1 | script | planner-pxqc-v1 | false | script | ok | script3b3 | script3b3/pl-school-008-r5-turn-2:solo | {
"pos": [
10,
8.8
],
"yaw_rad": -0.31,
"how": "near"
} | {
"minX": -1.704,
"maxX": 17.2,
"minZ": -0.679,
"maxZ": 16.3
} | false | method_v2 | false | 40 | 0 | {
"move": {
"name": "move",
"description": "Walk in a straight line. Stops early if something blocks the way.",
"parameters": {
"type": "object",
"properties": {
"distance_m": {
"type": "number",
"minimum": 0.3,
"maximum": 4,
"description": "metr... | [
{
"role": "system",
"text": "You are a game QA tester controlling a first-person agent in a 3D world through tools. Your job is to inspect the world for BUGS - anything broken, impossible, or inconsistent with how a normal game world should look and behave - and to report them precisely.\n\nStandard game-bu... | {
"task": "pl-school-008-r5-turn-2",
"arm": "solo",
"code": "T2",
"family": "turn",
"profile": "env",
"capturable": true,
"reported": 1,
"flags_total": 1,
"retracted": 0,
"success_lenient": true,
"success": true,
"success_clean": true,
"extra_reports": 0,
"first_flag_action": 21,
"n_action... | script:script3_success_clean.jsonl | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
pl-house-003-r7-upside-0~v1:solo:neg1h | pl-house-003-r7-upside-0 | solo | pl-house-003 | upside | S2 | 40 | env | mcp-v1 | azure | gpt-5.6-terra | false | neg | null | h100n1 | h100n1/pl-house-003-r7-upside-0~v1:solo:neg1h | {
"pos": [
0.2,
23
],
"yaw_rad": 2.297,
"how": "mid",
"near_share": 0.5
} | {
"minX": -5.7395,
"maxX": 8.35,
"minZ": 12.563,
"maxZ": 27.2
} | false | method_v2 | false | 36 | 39 | {"move":{"name":"move","description":"Walk in a straight line. Stops early if something blocks the w(...TRUNCATED) | [{"role":"system","text":"You are a game QA tester controlling a first-person agent in a 3D world th(...TRUNCATED) | {"task":"pl-house-003-r7-upside-0","arm":"solo","code":"S2","family":"upside","profile":"env","captu(...TRUNCATED) | neg1h | {"task":"pl-house-003-r7-upside-0","arm":"solo","code":"S2","family":"upside","profile":"env","captu(...TRUNCATED) | true | h100n1/pl-house-003-r7-upside-0~v1:solo | teacher | {"status":"confirmed","category":"semantic","evidence":["a29#c1","a35#c1"],"description":"Improper c(...TRUNCATED) | 39 | turn | [
"a29#c1",
"a35#c1"
] | null | null | model | azure:gpt-5.6-luna | 1 | {"provider":"azure","model":"gpt-5.6-terra","reasoning":"medium","max_out":4000,"temperature":null,"(...TRUNCATED) | null | null | null |
pl-street-009-r4-xray-0:solo | pl-street-009-r4-xray-0 | solo | pl-street-009 | xray | V1 | 40 | cottage | mcp-v1 | script | planner-pxqc-v1 | false | script | ok | script3b2 | script3b2/pl-street-009-r4-xray-0:solo | {
"pos": [
1.7000000000000002,
5.05
],
"yaw_rad": 2.665,
"how": "mid"
} | null | false | method_v2 | false | 8 | 0 | {"move":{"name":"move","description":"Walk in a straight line. Stops early if something blocks the w(...TRUNCATED) | [{"role":"system","text":"You are a game QA tester controlling a first-person agent in a 3D world th(...TRUNCATED) | {"task":"pl-street-009-r4-xray-0","arm":"solo","code":"V1","family":"xray","profile":"cottage","capt(...TRUNCATED) | script:script3_success_clean.jsonl | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
pl-street-004-r4-xray-5:solo | pl-street-004-r4-xray-5 | solo | pl-street-004 | xray | V1 | 40 | env | mcp-v1 | script | planner-pxqc-v1 | false | script | ok | script3b1 | script3b1/pl-street-004-r4-xray-5:solo | {
"pos": [
-2.7,
6.85
],
"yaw_rad": 1.845,
"how": "mid"
} | {
"minX": -21.7,
"maxX": -0.308,
"minZ": -3.546,
"maxZ": 20.454
} | false | method_v2 | false | 40 | 0 | {"move":{"name":"move","description":"Walk in a straight line. Stops early if something blocks the w(...TRUNCATED) | [{"role":"system","text":"You are a game QA tester controlling a first-person agent in a 3D world th(...TRUNCATED) | {"task":"pl-street-004-r4-xray-5","arm":"solo","code":"V1","family":"xray","profile":"env","capturab(...TRUNCATED) | script:script3_success_clean.jsonl | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
pl-school-002-r6-viewscale-0~v2:solo:neg1h | pl-school-002-r6-viewscale-0 | solo | pl-school-002 | viewscale | V2 | 40 | cottage | mcp-v1 | openai | qwen3.8-27b | false | neg | null | h100n2 | h100n2/pl-school-002-r6-viewscale-0~v2:solo:neg1h | {
"pos": [
-3.8,
14.4
],
"yaw_rad": 1.917,
"how": "mid",
"near_share": 0.5
} | {
"minX": -17.2,
"maxX": 3.753,
"minZ": 0.57,
"maxZ": 16.3
} | false | method_v2 | false | 8 | 14 | {"move":{"name":"move","description":"Walk in a straight line. Stops early if something blocks the w(...TRUNCATED) | [{"role":"system","text":"You are a game QA tester controlling a first-person agent in a 3D world th(...TRUNCATED) | {"task":"pl-school-002-r6-viewscale-0","arm":"solo","code":"V2","family":"viewscale","profile":"cott(...TRUNCATED) | neg1h | {"task":"pl-school-002-r6-viewscale-0","arm":"solo","code":"V2","family":"viewscale","profile":"cott(...TRUNCATED) | false | h100n2/pl-school-002-r6-viewscale-0~v2:solo | teacher | {"status":"confirmed","category":"visual","evidence":["a6","a7","a8"],"description":"Distance-depend(...TRUNCATED) | 14 | turn | [] | null | null | model | azure:gpt-5.6-luna | 2 | {"provider":"openai","model":"qwen3.8-27b","reasoning":"medium","max_out":16000,"temperature":1.0,"p(...TRUNCATED) | null | null | null |
pl-house-003-r4-tilt-0:solo | pl-house-003-r4-tilt-0 | solo | pl-house-003 | tilt | G2 | 40 | wilds | mcp-v1 | script | planner-pxqc-v1 | false | script | ok | script3b14 | script3b14/pl-house-003-r4-tilt-0:solo | {
"pos": [
-4.6,
1.8
],
"yaw_rad": 1.152,
"how": "mid"
} | {
"minX": -8.35,
"maxX": 8.35,
"minZ": -2.9,
"maxZ": 19.569
} | false | method_v2 | false | 40 | 0 | {"move":{"name":"move","description":"Walk in a straight line. Stops early if something blocks the w(...TRUNCATED) | [{"role":"system","text":"You are a game QA tester controlling a first-person agent in a 3D world th(...TRUNCATED) | {"task":"pl-house-003-r4-tilt-0","arm":"solo","code":"G2","family":"tilt","profile":"wilds","captura(...TRUNCATED) | script:script3_success_clean.jsonl | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
pl-house-007-r4-nosolid-3:solo | pl-house-007-r4-nosolid-3 | solo | pl-house-007 | nosolid | C1 | 40 | env | mcp-v1 | script | planner-pxqc-v1 | false | script | ok | script3b4 | script3b4/pl-house-007-r4-nosolid-3:solo | {
"pos": [
-2,
10.8
],
"yaw_rad": -0.031,
"how": "near"
} | {
"minX": -8.75,
"maxX": 8.75,
"minZ": 0.544,
"maxZ": 24.2
} | false | method_v2 | false | 29 | 0 | {"move":{"name":"move","description":"Walk in a straight line. Stops early if something blocks the w(...TRUNCATED) | [{"role":"system","text":"You are a game QA tester controlling a first-person agent in a 3D world th(...TRUNCATED) | {"task":"pl-house-007-r4-nosolid-3","arm":"solo","code":"C1","family":"nosolid","profile":"env","cap(...TRUNCATED) | script:script3_success_clean.jsonl | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
End of preview.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
VLMAuditor scripted trajectories, generator v2.1 (script3)
Synthetic training trajectories for small-VLM auditors in WorldAuditBench's MCP agent setting, generated on our own procedurally built worlds (not benchmark scenes).
script3_success_clean.jsonl: 7,806 records (formatmcp-v1:tools+messages), scripted explorer v2.1 (MCPGEN_SCRIPT_V=2.1), every record a clean success with its thoughts written; the 7 held-out evaluation worlds are excluded.images-script3-000..004.tar: the 49,829 images these records need that are not in the earlier (private) snapshots ofkimperyang/vlmauditor-mcpgen-data. Paths inside the tars are repo-relative (out/mcpgen-20261023/...): extract into the same IMAGE_ROOT as those snapshots.SHA256SUMS:sha256sum -c SHA256SUMS.
No benchmark tasks, rubrics or answers are included.
- Downloads last month
- 1,529