activitydiary JEV step1521 Add Activity
run_2b38f68437b6480c8f26d01cec61b440 · 第43.64分钟 · 请求step1521 / 决策step1521 / 执行step1521 · REPEATED_SUPPORTED · 证据MEDIUM · 结构化审查
一、模型为什么选择它
Native按exploratory-top3-v1聚合：没有confidence>0.5的SAME，存在可信DIFFERENT（p1 对step433=0.980000；p2 对step176=1.000000；p3 对step227=1.000000），因此PROMOTE。模型未返回自然语言理由；这里解释可复核的判断与聚合，不推测模型内心原因。
当前审核范围 g11s38 / g11a627；合法历史13条，比较3条。Pair缓存命中：False。
| Pair / 历史step | 原Evidence控件 | 原回答 | confidence | 可信阈值 |
|---|---|---|---|---|
| p1 / 433 | Select Activity | DIFFERENT_OPERATION | 0.98000000 | True |
| p2 / 176 | Map | DIFFERENT_OPERATION | 1.00000000 | True |
| p3 / 227 | Settings | DIFFERENT_OPERATION | 1.00000000 | True |
真正发送的state/questions · 原请求回执 · 原回答回执 · 本地审计（没有自动送模型）
展开原始Pair问题、回答和Witness字段
[
{
"pair_id": "p1",
"history_id": "h4",
"history_step": 433,
"history_input_id": "input-2369",
"choice": "DIFFERENT_OPERATION",
"confidence": 0.98,
"probabilities": {
"SAME_OPERATION": 0.01,
"DIFFERENT_OPERATION": 0.99,
"INSUFFICIENT_EVIDENCE": 0
},
"trusted": true,
"pair_cache_hit": false,
"original_question": {
"type": "choice",
"criteria": {
"SAME_OPERATION": "Same application-function operation; only parameter, data object, or entry point differs.",
"DIFFERENT_OPERATION": "Different application-function operations.",
"INSUFFICIENT_EVIDENCE": "The current visible GUI evidence is insufficient to determine the operation relation reliably."
},
"instructions": "Judge whether two concrete GUI actions represent the same application-function operation, using only the supplied GUI evidence and page context.\nSAME_OPERATION: the same application function, differing only in parameter, data object, or entry point.\nDIFFERENT_OPERATION: different application functions.\nINSUFFICIENT_EVIDENCE: the visible evidence is insufficient to judge reliably.\nThe same handler, class, resource ID, menu container, or switch is not sufficient evidence of operation equivalence. Judge application functions, not implementation form.\n\nIs candidate at `candidates.c1` the same application-function operation as executed action at `executed_actions.h4`? Interpret each action using its own source_page_id in pages.\n\nDecision standard:\nSAME_OPERATION means the candidate and the executed history are the same GUI operation pattern already represented; differences limited to data instances, entered values, or equivalent entry points are SAME.\nDIFFERENT_OPERATION means the candidate is a distinct command, feature entry, or interaction branch that the executed history does not represent and is reasonably worth one additional concrete-action attempt. Do not require proof of new method coverage, a new state, or successful completion.\nINSUFFICIENT_EVIDENCE means the supplied GUI and execution history do not support a reliable relation.\nTreat a visible menu-command label as evidence of the command. Do not equate a menu command with ordinary text entered into an EditText. Different Checklist rows with the same edit behavior are normally SAME. For a long-pressed EditText, do not infer a different block operation from the block name alone; it may be ordinary text selection. Similar class, resource ID, outer container, or ModelAction does not establish equivalence. Compare operation behavior, not wording alone."
},
"original_answer": {
"type": "choice",
"choice": "DIFFERENT_OPERATION",
"probabilities": {
"SAME_OPERATION": 0.01,
"DIFFERENT_OPERATION": 0.99,
"INSUFFICIENT_EVIDENCE": 0
},
"confidence": 0.98
},
"model_visible_history": {
"class": "android.widget.CheckedTextView",
"resource_id": "de.rampro.activitydiary:id/design_menu_item_text",
"text": "Select Activity",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": true,
"selected": false,
"action_type": "MODEL_CLICK",
"index": 1,
"concrete_identity": "MODEL_CLICK|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.drawerlayout.widget.DrawerLayout\"][@resource-id=\"de.rampro.activitydiary:id/drawer_layout\"][@text=\"\"][@content-desc=\"\"][@index=5][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"de.rampro.activitydiary:id/navigation_view\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.recyclerview.widget.RecyclerView\"][@resource-id=\"de.rampro.activitydiary:id/design_navigation_view\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='true']/*[@class=\"androidx.appcompat.widget.LinearLayoutCompat\"][@resource-id=\"de.rampro.activitydiary:id/nav_main\"][@text=\"\"][@content-desc=\"\"][@index=1][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.CheckedTextView\"][@resource-id=\"de.rampro.activitydiary:id/design_menu_item_text\"][@text=\"Select A\"][@content-desc=\"\"][@index=1][@enabled='true'][@clickable='true'][@checkable='true'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='true'][@long-clickable='false'][@scrollable='false']",
"model_action_id": "g11a627",
"source_state_id": "g11s38",
"bounds": [
42,
578,
693,
704
],
"source_page_id": "page5",
"target_subtree": [
{
"class": "android.widget.CheckedTextView",
"resource_id": "de.rampro.activitydiary:id/design_menu_item_text",
"text": "Select Activity",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": true,
"selected": false,
"path": []
}
],
"subtree_truncated": false,
"parent": {
"class": "androidx.appcompat.widget.LinearLayoutCompat",
"resource_id": "de.rampro.activitydiary:id/nav_main",
"text": "",
"content_desc": "",
"enabled": true,
"clickable": false,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false
},
"local_context": {
"parent": {
"class": "androidx.appcompat.widget.LinearLayoutCompat",
"resource_id": "de.rampro.activitydiary:id/nav_main"
},
"labels": [],
"truncated": false
},
"text_source": "retained_gui_text",
"generated_image_text": false,
"snapshot_provenance": "retained source GUI node; no inferred or reconstructed labels"
},
"offline_only_provenance": {
"source_tree": 515,
"source_gui_timestamp": 433,
"current_state": "g11s38",
"current_model_action": "g11a627",
"current_version": 33,
"input_confirmed": true,
"confirmation_basis": "live confirmedInputs before review; not function completion",
"attributes_recoverable": true
}
},
{
"pair_id": "p2",
"history_id": "h1",
"history_step": 176,
"history_input_id": "input-958",
"choice": "DIFFERENT_OPERATION",
"confidence": 1,
"probabilities": {
"SAME_OPERATION": 0,
"DIFFERENT_OPERATION": 1,
"INSUFFICIENT_EVIDENCE": 0
},
"trusted": true,
"pair_cache_hit": false,
"original_question": {
"type": "choice",
"criteria": {
"SAME_OPERATION": "Same application-function operation; only parameter, data object, or entry point differs.",
"DIFFERENT_OPERATION": "Different application-function operations.",
"INSUFFICIENT_EVIDENCE": "The current visible GUI evidence is insufficient to determine the operation relation reliably."
},
"instructions": "Judge whether two concrete GUI actions represent the same application-function operation, using only the supplied GUI evidence and page context.\nSAME_OPERATION: the same application function, differing only in parameter, data object, or entry point.\nDIFFERENT_OPERATION: different application functions.\nINSUFFICIENT_EVIDENCE: the visible evidence is insufficient to judge reliably.\nThe same handler, class, resource ID, menu container, or switch is not sufficient evidence of operation equivalence. Judge application functions, not implementation form.\n\nIs candidate at `candidates.c1` the same application-function operation as executed action at `executed_actions.h1`? Interpret each action using its own source_page_id in pages.\n\nDecision standard:\nSAME_OPERATION means the candidate and the executed history are the same GUI operation pattern already represented; differences limited to data instances, entered values, or equivalent entry points are SAME.\nDIFFERENT_OPERATION means the candidate is a distinct command, feature entry, or interaction branch that the executed history does not represent and is reasonably worth one additional concrete-action attempt. Do not require proof of new method coverage, a new state, or successful completion.\nINSUFFICIENT_EVIDENCE means the supplied GUI and execution history do not support a reliable relation.\nTreat a visible menu-command label as evidence of the command. Do not equate a menu command with ordinary text entered into an EditText. Different Checklist rows with the same edit behavior are normally SAME. For a long-pressed EditText, do not infer a different block operation from the block name alone; it may be ordinary text selection. Similar class, resource ID, outer container, or ModelAction does not establish equivalence. Compare operation behavior, not wording alone."
},
"original_answer": {
"type": "choice",
"choice": "DIFFERENT_OPERATION",
"probabilities": {
"SAME_OPERATION": 0,
"DIFFERENT_OPERATION": 1,
"INSUFFICIENT_EVIDENCE": 0
},
"confidence": 1
},
"model_visible_history": {
"class": "android.widget.CheckedTextView",
"resource_id": "de.rampro.activitydiary:id/design_menu_item_text",
"text": "Map",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false,
"action_type": "MODEL_CLICK",
"index": 4,
"concrete_identity": "MODEL_CLICK|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.drawerlayout.widget.DrawerLayout\"][@resource-id=\"de.rampro.activitydiary:id/drawer_layout\"][@text=\"\"][@content-desc=\"\"][@index=5][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"de.rampro.activitydiary:id/navigation_view\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.recyclerview.widget.RecyclerView\"][@resource-id=\"de.rampro.activitydiary:id/design_navigation_view\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='true']/*[@class=\"androidx.appcompat.widget.LinearLayoutCompat\"][@resource-id=\"de.rampro.activitydiary:id/nav_map\"][@text=\"\"][@content-desc=\"\"][@index=4][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.CheckedTextView\"][@resource-id=\"de.rampro.activitydiary:id/design_menu_item_text\"][@text=\"Map\"][@content-desc=\"\"][@index=4][@enabled='true'][@clickable='true'][@checkable='true'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='true'][@long-clickable='false'][@scrollable='false']",
"model_action_id": "g11a627",
"source_state_id": "g11s38",
"bounds": [
42,
956,
693,
1017
],
"source_page_id": "page2",
"target_subtree": [
{
"class": "android.widget.CheckedTextView",
"resource_id": "de.rampro.activitydiary:id/design_menu_item_text",
"text": "Map",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false,
"path": []
}
],
"subtree_truncated": false,
"parent": {
"class": "androidx.appcompat.widget.LinearLayoutCompat",
"resource_id": "de.rampro.activitydiary:id/nav_map",
"text": "",
"content_desc": "",
"enabled": true,
"clickable": false,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false
},
"local_context": {
"parent": {
"class": "androidx.appcompat.widget.LinearLayoutCompat",
"resource_id": "de.rampro.activitydiary:id/nav_map"
},
"labels": [],
"truncated": false
},
"text_source": "retained_gui_text",
"generated_image_text": false,
"snapshot_provenance": "retained source GUI node; no inferred or reconstructed labels"
},
"offline_only_provenance": {
"source_tree": 206,
"source_gui_timestamp": 176,
"current_state": "g11s38",
"current_model_action": "g11a627",
"current_version": 33,
"input_confirmed": true,
"confirmation_basis": "live confirmedInputs before review; not function completion",
"attributes_recoverable": true
}
},
{
"pair_id": "p3",
"history_id": "h2",
"history_step": 227,
"history_input_id": "input-1228",
"choice": "DIFFERENT_OPERATION",
"confidence": 1,
"probabilities": {
"SAME_OPERATION": 0,
"DIFFERENT_OPERATION": 1,
"INSUFFICIENT_EVIDENCE": 0
},
"trusted": true,
"pair_cache_hit": false,
"original_question": {
"type": "choice",
"criteria": {
"SAME_OPERATION": "Same application-function operation; only parameter, data object, or entry point differs.",
"DIFFERENT_OPERATION": "Different application-function operations.",
"INSUFFICIENT_EVIDENCE": "The current visible GUI evidence is insufficient to determine the operation relation reliably."
},
"instructions": "Judge whether two concrete GUI actions represent the same application-function operation, using only the supplied GUI evidence and page context.\nSAME_OPERATION: the same application function, differing only in parameter, data object, or entry point.\nDIFFERENT_OPERATION: different application functions.\nINSUFFICIENT_EVIDENCE: the visible evidence is insufficient to judge reliably.\nThe same handler, class, resource ID, menu container, or switch is not sufficient evidence of operation equivalence. Judge application functions, not implementation form.\n\nIs candidate at `candidates.c1` the same application-function operation as executed action at `executed_actions.h2`? Interpret each action using its own source_page_id in pages.\n\nDecision standard:\nSAME_OPERATION means the candidate and the executed history are the same GUI operation pattern already represented; differences limited to data instances, entered values, or equivalent entry points are SAME.\nDIFFERENT_OPERATION means the candidate is a distinct command, feature entry, or interaction branch that the executed history does not represent and is reasonably worth one additional concrete-action attempt. Do not require proof of new method coverage, a new state, or successful completion.\nINSUFFICIENT_EVIDENCE means the supplied GUI and execution history do not support a reliable relation.\nTreat a visible menu-command label as evidence of the command. Do not equate a menu command with ordinary text entered into an EditText. Different Checklist rows with the same edit behavior are normally SAME. For a long-pressed EditText, do not infer a different block operation from the block name alone; it may be ordinary text selection. Similar class, resource ID, outer container, or ModelAction does not establish equivalence. Compare operation behavior, not wording alone."
},
"original_answer": {
"type": "choice",
"choice": "DIFFERENT_OPERATION",
"probabilities": {
"SAME_OPERATION": 0,
"DIFFERENT_OPERATION": 1,
"INSUFFICIENT_EVIDENCE": 0
},
"confidence": 1
},
"model_visible_history": {
"class": "android.widget.CheckedTextView",
"resource_id": "de.rampro.activitydiary:id/design_menu_item_text",
"text": "Settings",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false,
"action_type": "MODEL_CLICK",
"index": 6,
"concrete_identity": "MODEL_CLICK|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.drawerlayout.widget.DrawerLayout\"][@resource-id=\"de.rampro.activitydiary:id/drawer_layout\"][@text=\"\"][@content-desc=\"\"][@index=5][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"de.rampro.activitydiary:id/navigation_view\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.recyclerview.widget.RecyclerView\"][@resource-id=\"de.rampro.activitydiary:id/design_navigation_view\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='true']/*[@class=\"androidx.appcompat.widget.LinearLayoutCompat\"][@resource-id=\"de.rampro.activitydiary:id/nav_settings\"][@text=\"\"][@content-desc=\"\"][@index=6][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.CheckedTextView\"][@resource-id=\"de.rampro.activitydiary:id/design_menu_item_text\"][@text=\"Settings\"][@content-desc=\"\"][@index=6][@enabled='true'][@clickable='true'][@checkable='true'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='true'][@long-clickable='false'][@scrollable='false']",
"model_action_id": "g11a627",
"source_state_id": "g11s38",
"bounds": [
42,
785,
693,
911
],
"source_page_id": "page3",
"target_subtree": [
{
"class": "android.widget.CheckedTextView",
"resource_id": "de.rampro.activitydiary:id/design_menu_item_text",
"text": "Settings",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false,
"path": []
}
],
"subtree_truncated": false,
"parent": {
"class": "androidx.appcompat.widget.LinearLayoutCompat",
"resource_id": "de.rampro.activitydiary:id/nav_settings",
"text": "",
"content_desc": "",
"enabled": true,
"clickable": false,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false
},
"local_context": {
"parent": {
"class": "androidx.appcompat.widget.LinearLayoutCompat",
"resource_id": "de.rampro.activitydiary:id/nav_settings"
},
"labels": [],
"truncated": false
},
"text_source": "retained_gui_text",
"generated_image_text": false,
"snapshot_provenance": "retained source GUI node; no inferred or reconstructed labels"
},
"offline_only_provenance": {
"source_tree": 272,
"source_gui_timestamp": 227,
"current_state": "g11s38",
"current_model_action": "g11a627",
"current_version": 33,
"input_confirmed": true,
"confirmation_basis": "live confirmedInputs before review; not function completion",
"attributes_recoverable": true
}
}
]二、此前APE实际做过什么
普通407/1029已尝试Add Activity并进入EditActivity；本次打开标题/颜色表单，没有保存新活动的证明。
等价历史是否送入Provider wire：NO；当前Pair直接比较过的等价历史step：[]。这是wire可见性，服务端最终token渲染是否逐字段保留仍UNKNOWN。
| 此前step / UTC | 页面与控件 | 确认与目标 |
|---|---|---|
| 407 2026-11-06T22:59:49.600+00:00 | Add Activity de.rampro.activitydiary.ui.main.MainActivity | Native确认=True de.rampro.activitydiary.ui.generic.EditActivity 保留目标GUI |
| 1029 2026-11-06T23:14:27.750+00:00 | Add Acti de.rampro.activitydiary.ui.main.MainActivity | Native确认=True de.rampro.activitydiary.ui.generic.EditActivity 目标GUI未保留，不推断业务完成 |
三、Probe执行后发生什么
Add Activity 打开 New activity 标题和颜色表单；没有保存结果证据。
Run内首次METHOD观察：完成后2秒=0，10秒=0，下一动作前B=0。仅时间关联；A2/A10可含后续动作。原方法完整名单及时间戳。
后续APE输入与生命周期原始事件
{
"following_actions": [
{
"input_id": "input-8369",
"step": 1522,
"version": 33,
"elapsed_s": 2620.051,
"wall_time_ms": 1794008012179,
"next_input_wall_ms": 1794008013428,
"input_completed_wall_ms": 1794008012204,
"input_confirmed": true,
"action": "MODEL_CLICK Save",
"action_type": "MODEL_CLICK",
"text": "",
"content_desc": "Save",
"resource_id": "de.rampro.activitydiary:id/action_edit_done",
"model_action_id": "g20a1553",
"concrete_identity": "MODEL_CLICK|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.TextView\"][@resource-id=\"de.rampro.activitydiary:id/action_edit_done\"][@text=\"\"][@content-desc=\"Save\"][@index=3][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false']",
"state_id": "g20s117",
"source_gui_id": 1646,
"activity": "de.rampro.activitydiary.ui.generic.EditActivity",
"activity_evidence": "GUI receipt",
"labels": [
"#c3462800",
"New activity",
"Delete activity",
"Save",
"Activity title",
"Activity color"
],
"probe": false,
"strategy": "EPSILON_GREEDY",
"context": "exploration",
"native_branch": "Epsilon",
"executor_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/executor.log:149492",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/native/sata-de.rampro.activitydiary-ape-sata-running-minutes-60/jev-audit/inputs/generated-8370.json"
},
{
"input_id": "input-8374",
"step": 1523,
"version": 33,
"elapsed_s": 2621.3,
"wall_time_ms": 1794008013428,
"next_input_wall_ms": 1794008013458,
"input_completed_wall_ms": 1794008013456,
"input_confirmed": true,
"action": "MODEL_BACK NA",
"action_type": "MODEL_BACK",
"text": null,
"content_desc": null,
"resource_id": null,
"model_action_id": "g7a387",
"concrete_identity": null,
"state_id": "g7s17",
"source_gui_id": 1647,
"activity": null,
"activity_evidence": null,
"labels": null,
"probe": false,
"strategy": "EARLY_STAGE",
"context": "exploration",
"native_branch": "Local Greedy",
"executor_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/executor.log:149576",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/native/sata-de.rampro.activitydiary-ape-sata-running-minutes-60/jev-audit/inputs/generated-8375.json"
},
{
"input_id": "other-8379",
"step": 1523,
"version": 33,
"elapsed_s": 2621.33,
"wall_time_ms": 1794008013458,
"next_input_wall_ms": 1794008018128,
"input_completed_wall_ms": null,
"input_confirmed": false,
"action": "EVENT_START EVENT_START",
"action_type": "EVENT_START",
"text": null,
"content_desc": null,
"resource_id": null,
"model_action_id": null,
"concrete_identity": null,
"state_id": "g7s17",
"source_gui_id": null,
"activity": null,
"activity_evidence": null,
"labels": null,
"probe": false,
"strategy": "EARLY_STAGE",
"context": "exploration",
"native_branch": "Local Greedy",
"executor_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/executor.log:149576",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/native/sata-de.rampro.activitydiary-ape-sata-running-minutes-60/jev-audit/inputs/other-8380.json"
},
{
"input_id": "input-8385",
"step": 1524,
"version": 33,
"elapsed_s": 2626.0,
"wall_time_ms": 1794008018128,
"next_input_wall_ms": 1794008019040,
"input_completed_wall_ms": 1794008018154,
"input_confirmed": true,
"action": "MODEL_CLICK Search",
"action_type": "MODEL_CLICK",
"text": "",
"content_desc": "Search",
"resource_id": "de.rampro.activitydiary:id/search_button",
"model_action_id": "g7a364",
"concrete_identity": "MODEL_CLICK|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.appcompat.widget.LinearLayoutCompat\"][@resource-id=\"de.rampro.activitydiary:id/action_filter\"][@text=\"\"][@content-desc=\"\"][@index=3][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.ImageView\"][@resource-id=\"de.rampro.activitydiary:id/search_button\"][@text=\"\"][@content-desc=\"Search\"][@index=0][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false']",
"state_id": "g7s17",
"source_gui_id": 1648,
"activity": "de.rampro.activitydiary.ui.main.MainActivity",
"activity_evidence": "GUI receipt",
"labels": [
"#8fb7fa14",
"Activity Diary",
"What are you doing?",
"Add Activity",
"Search",
"#793ec2d9",
"#28bf6eaa",
"Statistics",
"Note",
"Pictures",
"STATISTICS",
"NOTE",
"PICTURES",
"Cooking",
"Relaxing",
"Woodworking",
"Cinema",
"Gardening",
"Sleeping",
"Cleaning",
"Officework",
"Swimming",
"Since just now",
"-",
"Sat, 11.07.26: 0 sec",
"2026 Week 45: 0 sec",
"November 2026: 0 sec"
],
"probe": false,
"strategy": "EARLY_STAGE",
"context": "navigation",
"native_branch": "Remote Greedy",
"executor_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/executor.log:149704",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/native/sata-de.rampro.activitydiary-ape-sata-running-minutes-60/jev-audit/inputs/generated-8386.json"
},
{
"input_id": "input-8390",
"step": 1525,
"version": 33,
"elapsed_s": 2626.912,
"wall_time_ms": 1794008019040,
"next_input_wall_ms": 1794008021653,
"input_completed_wall_ms": 1794008020523,
"input_confirmed": true,
"action": "MODEL_CLICK Search…",
"action_type": "MODEL_CLICK",
"text": "Search…",
"content_desc": "",
"resource_id": "de.rampro.activitydiary:id/search_src_text",
"model_action_id": "g13a703",
"concrete_identity": "MODEL_CLICK|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.appcompat.widget.LinearLayoutCompat\"][@resource-id=\"de.rampro.activitydiary:id/action_filter\"][@text=\"\"][@content-desc=\"\"][@index=3][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.AutoCompleteTextView\"][@resource-id=\"de.rampro.activitydiary:id/search_src_text\"][@text=\"Search…\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='true'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='true'][@scrollable='false']",
"state_id": "g13s50",
"source_gui_id": 1649,
"activity": "de.rampro.activitydiary.ui.main.MainActivity",
"activity_evidence": "exact source_gui_id transition receipt",
"labels": null,
"probe": false,
"strategy": "USE_BUFFER",
"context": "exploration",
"native_branch": "USE_BUFFER",
"executor_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/executor.log:149821",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_2b38f68437b6480c8f26d01cec61b440.zip!device-output/native/sata-de.rampro.activitydiary-ape-sata-running-minutes-60/jev-audit/inputs/generated-8392.json"
}
],
"lifecycle": {
"DECISION_PROBE": {
"event": "DECISION_PROBE",
"step": 1521,
"monotonic_ns": 1897058245456173,
"wall_time_ms": 1794008010494,
"abstraction_version": 33,
"state_id": "g11s38",
"model_action_id": "g11a627",
"entry_id": "semantic-172-c1",
"request_id": "semantic-172",
"candidate_id": "c1",
"confidence": 1,
"choice": "PROMOTE",
"parsed_outcome": "PROBE",
"ledger_status": "STAGED",
"decision_role": "candidate_decision",
"input_signature": "ac114de916d53fda86ef84b5a7a37f590978f52abf5845574616db31671e1702",
"concrete_identity": "7f21594c87b4d4db48a98c88eb9ee350f44e668c540806fd79bb7e8127ee2461",
"decision_scope": "396bcc4926a1bbc717e30d3a206a093f7d70264dcb5dc1ffbcaa231e8afc1956",
"representative_model_action_id": "PROMOTE",
"decision_request_id": "semantic-172",
"aggregation_policy": "exploratory-top3-v1",
"total_history": 13,
"history_complete": true,
"unknown_mapping_count": 0,
"recoverable_history_count": 13,
"selected_history": 3,
"comparison_count": 3,
"same_count": 0,
"different_count": 3,
"insufficient_count": 0,
"trusted_same_count": 0,
"trusted_different_count": 3,
"low_confidence_count": 0,
"full_history_compared": false,
"exploratory": true,
"_line": 6479
},
"SEMANTIC_DECISION_STORED": {
"event": "SEMANTIC_DECISION_STORED",
"step": 1521,
"monotonic_ns": 1897058246187172,
"wall_time_ms": 1794008010495,
"abstraction_version": 33,
"state_id": "g11s38",
"model_action_id": "g11a627",
"entry_id": "semantic-172-c1",
"request_id": "semantic-172",
"candidate_id": "c1",
"decision_role": "candidate_decision",
"ledger_status": "PROBE",
"input_signature": "ac114de916d53fda86ef84b5a7a37f590978f52abf5845574616db31671e1702",
"input_fingerprint": "e06a066743d1b135ff63cc3c78e31292c9b4e35990c04fa685aa5278bed3db3a",
"concrete_identity": "7f21594c87b4d4db48a98c88eb9ee350f44e668c540806fd79bb7e8127ee2461",
"decision_scope": "396bcc4926a1bbc717e30d3a206a093f7d70264dcb5dc1ffbcaa231e8afc1956",
"_line": 6480
},
"PROBE_BOUND": {
"event": "PROBE_BOUND",
"step": 1521,
"monotonic_ns": 1897058278381888,
"wall_time_ms": 1794008010527,
"abstraction_version": 33,
"state_id": "g11s38",
"model_action_id": "g11a627",
"entry_id": "semantic-172-c1",
"request_id": "semantic-172",
"candidate_id": "c1",
"confidence": 1,
"choice": "PROMOTE",
"source": "semantic_sync",
"input_signature": "ac114de916d53fda86ef84b5a7a37f590978f52abf5845574616db31671e1702",
"decision_scope": "396bcc4926a1bbc717e30d3a206a093f7d70264dcb5dc1ffbcaa231e8afc1956",
"concrete_identity": "7f21594c87b4d4db48a98c88eb9ee350f44e668c540806fd79bb7e8127ee2461",
"_line": 6481
},
"PROBE_SELECTED": {
"event": "PROBE_SELECTED",
"step": 1521,
"monotonic_ns": 1897058278414217,
"wall_time_ms": 1794008010527,
"abstraction_version": 33,
"state_id": "g11s38",
"model_action_id": "g11a627",
"entry_id": "semantic-172-c1",
"request_id": "semantic-172",
"candidate_id": "c1",
"confidence": 1,
"choice": "PROMOTE",
"source": "semantic_sync",
"_line": 6482
},
"PROBE_INPUT_GENERATED": {
"event": "PROBE_INPUT_GENERATED",
"step": 1521,
"monotonic_ns": 1897058320299897,
"wall_time_ms": 1794008010569,
"abstraction_version": 33,
"state_id": null,
"model_action_id": null,
"entry_id": "semantic-172-c1",
"request_id": "semantic-172",
"candidate_id": "c1",
"event_count": 2,
"_line": 6485
},
"PROBE_INPUT_COMPLETED": {
"event": "PROBE_INPUT_COMPLETED",
"step": 1521,
"monotonic_ns": 1897058356552737,
"wall_time_ms": 1794008010605,
"abstraction_version": 33,
"state_id": null,
"model_action_id": null,
"input_phase": "input_injected",
"event_count": 2,
"injected_success": 2,
"injected_failed": 0,
"discarded": 0,
"unreported": 0,
"entry_id": "semantic-172-c1",
"request_id": "semantic-172",
"candidate_id": "c1",
"_line": 6486
},
"PROBE_EXECUTED": {
"event": "PROBE_EXECUTED",
"step": 1522,
"monotonic_ns": 1897059619808722,
"wall_time_ms": 1794008011868,
"abstraction_version": 33,
"state_id": "g11s38",
"model_action_id": "g11a627",
"entry_id": "semantic-172-c1",
"request_id": "semantic-172",
"candidate_id": "c1",
"confidence": 1,
"choice": "PROMOTE",
"transition_id": "g33e1049:1500",
"target_state": "g20s117",
"input_confirmed": true,
"_line": 6488
}
},
"selection_at_execution": {
"event": "SELECTION_DECISION",
"step": 1521,
"monotonic_ns": 1897058278645677,
"wall_time_ms": 1794008010527,
"abstraction_version": 33,
"state_id": "g11s38",
"model_action_id": "g11a627",
"snapshot_reason": "selection",
"state_visits": 40,
"state_first_seen": 103,
"state_last_seen": 1521,
"state_saturated": true,
"state_saturation": 1,
"native_state_saturated": true,
"native_unsaturated_model_action_count": 0,
"model_action_count": 9,
"enabled_valid_action_count": 9,
"unsaturated_model_action_count": 0,
"unvisited_model_action_count": 0,
"effective_unvisited_model_action_count": 0,
"effective_state_saturated": true,
"local_probe_action_count": 0,
"navigation_action_count": 0,
"enabled_valid_unvisited_action_count": 0,
"unvisited_model_action_ids": "",
"selection_strategy": "PROBE",
"selection_context": "exploration",
"selection_buffer_size_after": 0,
"selected_model_action_type": "MODEL_CLICK",
"selected_model_action_saturated": true,
"selected_model_action_unvisited": false,
"selected_effective_unvisited": false,
"selected_navigation_target": false,
"selected_model_action_visits": 15,
"selected_action_fingerprint": "type=MODEL_CLICK|activity=de.rampro.activitydiary.ui.main.MainActivity|class=android.widget.CheckedTextView|text=Add Acti|resource_id=de.rampro.activitydiary:id/design_menu_item_text|bounds=42,989,693,1017",
"selected_resolved_node_count": 5,
"_line": 6484
}
}四、是否值得PROMOTE
完整已保存历史支持此前已尝试等价功能；不能自动归为模型错误。
疑似主要环节：HISTORY_SCOPE；wire关系审查：UNKNOWN。
冻结实现：Registry历史范围（315–342） · Top3（210–230）及聚合（146–178） · Ledger pending与确认（235–285） · Probe调度机会（415–468）。源码行号是原文件行号，GUI历史与源码解释分别有边界。
即便新功能也只证明此次Probe填补了该Run当时的遗漏；正常APE仍可经后续Epsilon等策略选到，不能推出Pure永远不会执行。全部事实与原始位置 · 本Run完整保留输入历史