wikipedia STARTLUX step450 Show all
run_d04ae609a94042e28f3ab00ddb284ca9 · 第12.46分钟 · 请求step450 / 决策step450 / 执行step450 · UNKNOWN · 证据INSUFFICIENT_FOR_GLOBAL_NOVELTY · 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step448=0.720124；p2 对step446=0.818250），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 g28s153 / g28a1671；合法历史2条，比较2条。Pair缓存命中：False。
| Pair / 历史step | 原Evidence控件 | 原回答 | confidence | 可信阈值 |
|---|---|---|---|---|
| p1 / 448 | Hide all | DIFFERENT_OPERATION | 0.72012436 | True |
| p2 / 446 | Restore default view | DIFFERENT_OPERATION | 0.81824985 | 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": "h2",
"history_step": 448,
"history_input_id": "input-2353",
"choice": "DIFFERENT_OPERATION",
"confidence": 0.720124363899231,
"probabilities": {
"SAME_OPERATION": 0.11688928306102753,
"DIFFERENT_OPERATION": 0.8134162425994873,
"INSUFFICIENT_EVIDENCE": 0.06969448179006577
},
"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",
"confidence": 0.720124363899231,
"probabilities": {
"SAME_OPERATION": 0.11688928306102753,
"DIFFERENT_OPERATION": 0.8134162425994873,
"INSUFFICIENT_EVIDENCE": 0.06969448179006577
}
},
"model_visible_history": {
"class": "android.widget.TextView",
"resource_id": "org.wikipedia:id/title",
"text": "Hide all",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"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=\"android.widget.ListView\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.LinearLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=1][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.TextView\"][@resource-id=\"org.wikipedia:id/title\"][@text=\"Hide all\"][@content-desc=\"\"][@index=1][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false']",
"model_action_id": "g28a1671",
"source_state_id": "g28s153",
"bounds": [
597,
265,
1028,
322
],
"source_page_id": "page3",
"target_subtree": [
{
"class": "android.widget.TextView",
"resource_id": "org.wikipedia:id/title",
"text": "Hide all",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false,
"path": []
}
],
"subtree_truncated": false,
"parent": {
"class": "android.widget.LinearLayout",
"resource_id": "",
"text": "",
"content_desc": "",
"enabled": true,
"clickable": false,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false
},
"local_context": {
"parent": {
"class": "android.widget.LinearLayout",
"resource_id": ""
},
"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": 594,
"source_gui_timestamp": 448,
"current_state": "g28s153",
"current_model_action": "g28a1671",
"current_version": 28,
"input_confirmed": true,
"confirmation_basis": "live confirmedInputs before review; not function completion",
"attributes_recoverable": true
}
},
{
"pair_id": "p2",
"history_id": "h1",
"history_step": 446,
"history_input_id": "input-2339",
"choice": "DIFFERENT_OPERATION",
"confidence": 0.8182498514652252,
"probabilities": {
"SAME_OPERATION": 0.06113607808947563,
"DIFFERENT_OPERATION": 0.8788332343101501,
"INSUFFICIENT_EVIDENCE": 0.06003072112798691
},
"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",
"confidence": 0.8182498514652252,
"probabilities": {
"SAME_OPERATION": 0.06113607808947563,
"DIFFERENT_OPERATION": 0.8788332343101501,
"INSUFFICIENT_EVIDENCE": 0.06003072112798691
}
},
"model_visible_history": {
"class": "android.widget.TextView",
"resource_id": "org.wikipedia:id/title",
"text": "Restore default view",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false,
"action_type": "MODEL_CLICK",
"index": 2,
"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.ListView\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.LinearLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=2][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.TextView\"][@resource-id=\"org.wikipedia:id/title\"][@text=\"Restore \"][@content-desc=\"\"][@index=2][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false']",
"model_action_id": "g28a1671",
"source_state_id": "g28s153",
"bounds": [
597,
391,
1028,
448
],
"source_page_id": "page2",
"target_subtree": [
{
"class": "android.widget.TextView",
"resource_id": "org.wikipedia:id/title",
"text": "Restore default view",
"content_desc": "",
"enabled": true,
"clickable": true,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false,
"path": []
}
],
"subtree_truncated": false,
"parent": {
"class": "android.widget.LinearLayout",
"resource_id": "",
"text": "",
"content_desc": "",
"enabled": true,
"clickable": false,
"long_clickable": false,
"scrollable": false,
"checked": false,
"selected": false
},
"local_context": {
"parent": {
"class": "android.widget.LinearLayout",
"resource_id": ""
},
"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": 592,
"source_gui_timestamp": 446,
"current_state": "g28s153",
"current_model_action": "g28a1671",
"current_version": 28,
"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
现有Witness支持或提示不同命令/操作分支，但没有足够完整的Run内等价历史恢复，不能从“未检索到”推出从未执行。保留UNKNOWN；不将新增方法或零新增单独作为PROMOTE对错。
等价历史是否送入Provider wire：UNKNOWN；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。
NA：没有足以可靠裁定全局首次或重复的已核实等价历史。不能把未检索到当成未执行。
三、Probe执行后发生什么
Show all 将Explore feed各开关全开；10秒627首次方法包含随后进入ReadingLists等普通操作影响。
Run内首次METHOD观察：完成后2秒=0，10秒=627，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-2377",
"step": 451,
"version": 28,
"elapsed_s": 748.553,
"wall_time_ms": 1793991237599,
"next_input_wall_ms": 1793991238886,
"input_completed_wall_ms": 1793991237969,
"input_confirmed": true,
"action": "MODEL_SCROLL_LEFT_RIGHT org.wikipedia:id/content_types_recycler",
"action_type": "MODEL_SCROLL_LEFT_RIGHT",
"text": "",
"content_desc": "",
"resource_id": "org.wikipedia:id/content_types_recycler",
"model_action_id": "g24a1573",
"concrete_identity": "MODEL_SCROLL_LEFT_RIGHT|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.recyclerview.widget.RecyclerView\"][@resource-id=\"org.wikipedia:id/content_types_recycler\"][@text=\"\"][@content-desc=\"\"][@index=3][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='true'][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='true']",
"state_id": "g24s145",
"source_gui_id": 597,
"activity": "org.wikipedia.feed.configure.ConfigureActivity",
"activity_evidence": "GUI receipt",
"labels": [
"#f1409233",
"Customize the feed",
"More options",
"Featured article, Daily featured article on Wikipedia",
"Top read, Daily most-viewed articles",
"Picture of the day, Daily featured image from Wikimedia Commons",
"Because you read, Suggestions based on a recently read article from your history",
"In the news, Articles about current events",
"On this day, Events in history on this day",
"Randomizer, Generate random articles to read",
"Today on Wikipedia, Main page of Wikipedia with daily featured content",
"Featured article",
"Daily featured article on Wikipedia",
"Top read",
"Daily most-viewed articles",
"Picture of the day",
"Daily featured image from Wikimedia Commons",
"Because you read",
"Suggestions based on a recently read article from your history",
"In the news",
"Articles about current events",
"On this day",
"Events in history on this day",
"Randomizer",
"Generate random articles to read",
"Today on Wikipedia",
"EN",
"IT",
"ZH-\nTW",
"ZH-\nCN",
"FR"
],
"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_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/executor.log:37683",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/native/sata-org.wikipedia-ape-sata-running-minutes-60/jev-audit/inputs/generated-2378.json"
},
{
"input_id": "input-2382",
"step": 452,
"version": 28,
"elapsed_s": 749.84,
"wall_time_ms": 1793991238886,
"next_input_wall_ms": 1793991240210,
"input_completed_wall_ms": 1793991239251,
"input_confirmed": true,
"action": "MODEL_SCROLL_TOP_DOWN org.wikipedia:id/content_types_recycler",
"action_type": "MODEL_SCROLL_TOP_DOWN",
"text": "",
"content_desc": "",
"resource_id": "org.wikipedia:id/content_types_recycler",
"model_action_id": "g24a1572",
"concrete_identity": "MODEL_SCROLL_TOP_DOWN|/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"androidx.recyclerview.widget.RecyclerView\"][@resource-id=\"org.wikipedia:id/content_types_recycler\"][@text=\"\"][@content-desc=\"\"][@index=3][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='true'][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='true']",
"state_id": "g24s145",
"source_gui_id": 598,
"activity": "org.wikipedia.feed.configure.ConfigureActivity",
"activity_evidence": "exact source_gui_id transition receipt",
"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_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/executor.log:37794",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/native/sata-org.wikipedia-ape-sata-running-minutes-60/jev-audit/inputs/generated-2383.json"
},
{
"input_id": "input-2387",
"step": 453,
"version": 28,
"elapsed_s": 751.164,
"wall_time_ms": 1793991240210,
"next_input_wall_ms": 1793991241151,
"input_completed_wall_ms": 1793991240235,
"input_confirmed": true,
"action": "MODEL_CLICK #f1409233",
"action_type": "MODEL_CLICK",
"text": "#f1409233",
"content_desc": "Navigate up",
"resource_id": "",
"model_action_id": "g24a1568",
"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.ImageButton\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"Navigate up\"][@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": "g24s145",
"source_gui_id": 599,
"activity": "org.wikipedia.feed.configure.ConfigureActivity",
"activity_evidence": "exact source_gui_id transition receipt",
"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_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/executor.log:37899",
"observed_methods_until_next": 13,
"category_counts": {
"app_domain": 9,
"third_party": 4
},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/native/sata-org.wikipedia-ape-sata-running-minutes-60/jev-audit/inputs/generated-2388.json"
},
{
"input_id": "input-2392",
"step": 454,
"version": 28,
"elapsed_s": 752.105,
"wall_time_ms": 1793991241151,
"next_input_wall_ms": 1793991242067,
"input_completed_wall_ms": 1793991241185,
"input_confirmed": true,
"action": "MODEL_CLICK More",
"action_type": "MODEL_CLICK",
"text": "",
"content_desc": "More",
"resource_id": "org.wikipedia:id/nav_more_container",
"model_action_id": "g28a1673",
"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.FrameLayout\"][@resource-id=\"org.wikipedia:id/nav_more_container\"][@text=\"\"][@content-desc=\"More\"][@index=2][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false']",
"state_id": "g28s154",
"source_gui_id": 600,
"activity": "org.wikipedia.main.MainActivity",
"activity_evidence": "exact source_gui_id transition receipt",
"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_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/executor.log:38005",
"observed_methods_until_next": 0,
"category_counts": {},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/native/sata-org.wikipedia-ape-sata-running-minutes-60/jev-audit/inputs/generated-2393.json"
},
{
"input_id": "input-2397",
"step": 455,
"version": 28,
"elapsed_s": 753.021,
"wall_time_ms": 1793991242067,
"next_input_wall_ms": 1793991243323,
"input_completed_wall_ms": 1793991242102,
"input_confirmed": true,
"action": "MODEL_CLICK Places",
"action_type": "MODEL_CLICK",
"text": "Places",
"content_desc": "",
"resource_id": "",
"model_action_id": "g8a974",
"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.view.ViewGroup\"][@resource-id=\"org.wikipedia:id/coordinator\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.FrameLayout\"][@resource-id=\"org.wikipedia:id/design_bottom_sheet\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.ScrollView\"][@resource-id=\"\"][@text=\"\"][@content-desc=\"\"][@index=0][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.LinearLayout\"][@resource-id=\"org.wikipedia:id/main_drawer_places_container\"][@text=\"\"][@content-desc=\"\"][@index=1][@enabled='true'][@clickable='false'][@checkable='false'][@long-clickable='false'][@scrollable='false']/*[@class=\"android.widget.TextView\"][@resource-id=\"\"][@text=\"Places\"][@content-desc=\"\"][@index=1][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false'][@enabled='true'][@clickable='true'][@checkable='false'][@long-clickable='false'][@scrollable='false']",
"state_id": "g8s93",
"source_gui_id": 601,
"activity": "org.wikipedia.main.MainActivity",
"activity_evidence": "GUI receipt",
"labels": [
"Log in / join Wikipedia",
"Places",
"Settings",
"Donate"
],
"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_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/executor.log:38081",
"observed_methods_until_next": 477,
"category_counts": {
"third_party": 466,
"app_domain": 11
},
"inventory_ref": "[WORKSPACE]/experiments/guitaboo-5app-1rep-baseline-comparison-20261107/analysis/evidence-zips/run_d04ae609a94042e28f3ab00ddb284ca9.zip!device-output/native/sata-org.wikipedia-ape-sata-running-minutes-60/jev-audit/inputs/generated-2398.json"
}
],
"lifecycle": {
"DECISION_PROBE": {
"event": "DECISION_PROBE",
"step": 450,
"monotonic_ns": 1880283575622616,
"wall_time_ms": 1793991236672,
"abstraction_version": 28,
"state_id": "g28s153",
"model_action_id": "g28a1671",
"entry_id": "semantic-4-c1",
"request_id": "semantic-4",
"candidate_id": "c1",
"confidence": 0.8182498514652252,
"choice": "PROMOTE",
"parsed_outcome": "PROBE",
"ledger_status": "STAGED",
"decision_role": "candidate_decision",
"input_signature": "5531b6a28314ad28f4814eb09bfbc2d012b1af3aa2e25955f12915503f5ee550",
"concrete_identity": "e2577dc7a52ff280dcb2a800bb721c622b505921c4bf70f9daebe581b410e387",
"decision_scope": "b77aa557953e35f2c783bdc341490f9b66c37955d1e455356229bcbfbafa56e1",
"representative_model_action_id": "PROMOTE",
"decision_request_id": "semantic-4",
"aggregation_policy": "exploratory-top3-v1",
"total_history": 2,
"history_complete": true,
"unknown_mapping_count": 0,
"recoverable_history_count": 2,
"selected_history": 2,
"comparison_count": 2,
"same_count": 0,
"different_count": 2,
"insufficient_count": 0,
"trusted_same_count": 0,
"trusted_different_count": 2,
"low_confidence_count": 0,
"full_history_compared": true,
"exploratory": false,
"_line": 1097
},
"SEMANTIC_DECISION_STORED": {
"event": "SEMANTIC_DECISION_STORED",
"step": 450,
"monotonic_ns": 1880283576193480,
"wall_time_ms": 1793991236672,
"abstraction_version": 28,
"state_id": "g28s153",
"model_action_id": "g28a1671",
"entry_id": "semantic-4-c1",
"request_id": "semantic-4",
"candidate_id": "c1",
"decision_role": "candidate_decision",
"ledger_status": "PROBE",
"input_signature": "5531b6a28314ad28f4814eb09bfbc2d012b1af3aa2e25955f12915503f5ee550",
"input_fingerprint": "47f60f19f8dd956a5ec11182ce0f1ed4f324613f998e9589dedb44d120ea3679",
"concrete_identity": "e2577dc7a52ff280dcb2a800bb721c622b505921c4bf70f9daebe581b410e387",
"decision_scope": "b77aa557953e35f2c783bdc341490f9b66c37955d1e455356229bcbfbafa56e1",
"_line": 1098
},
"PROBE_BOUND": {
"event": "PROBE_BOUND",
"step": 450,
"monotonic_ns": 1880283582670783,
"wall_time_ms": 1793991236679,
"abstraction_version": 28,
"state_id": "g28s153",
"model_action_id": "g28a1671",
"entry_id": "semantic-4-c1",
"request_id": "semantic-4",
"candidate_id": "c1",
"confidence": 0.8182498514652252,
"choice": "PROMOTE",
"source": "semantic_sync",
"input_signature": "5531b6a28314ad28f4814eb09bfbc2d012b1af3aa2e25955f12915503f5ee550",
"decision_scope": "b77aa557953e35f2c783bdc341490f9b66c37955d1e455356229bcbfbafa56e1",
"concrete_identity": "e2577dc7a52ff280dcb2a800bb721c622b505921c4bf70f9daebe581b410e387",
"_line": 1099
},
"PROBE_SELECTED": {
"event": "PROBE_SELECTED",
"step": 450,
"monotonic_ns": 1880283582726089,
"wall_time_ms": 1793991236679,
"abstraction_version": 28,
"state_id": "g28s153",
"model_action_id": "g28a1671",
"entry_id": "semantic-4-c1",
"request_id": "semantic-4",
"candidate_id": "c1",
"confidence": 0.8182498514652252,
"choice": "PROMOTE",
"source": "semantic_sync",
"_line": 1100
},
"PROBE_INPUT_GENERATED": {
"event": "PROBE_INPUT_GENERATED",
"step": 450,
"monotonic_ns": 1880283634610400,
"wall_time_ms": 1793991236731,
"abstraction_version": 28,
"state_id": null,
"model_action_id": null,
"entry_id": "semantic-4-c1",
"request_id": "semantic-4",
"candidate_id": "c1",
"event_count": 2,
"_line": 1103
},
"PROBE_INPUT_COMPLETED": {
"event": "PROBE_INPUT_COMPLETED",
"step": 450,
"monotonic_ns": 1880283672076372,
"wall_time_ms": 1793991236768,
"abstraction_version": 28,
"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-4-c1",
"request_id": "semantic-4",
"candidate_id": "c1",
"_line": 1104
},
"PROBE_EXECUTED": {
"event": "PROBE_EXECUTED",
"step": 451,
"monotonic_ns": 1880284495936022,
"wall_time_ms": 1793991237592,
"abstraction_version": 28,
"state_id": "g28s153",
"model_action_id": "g28a1671",
"entry_id": "semantic-4-c1",
"request_id": "semantic-4",
"candidate_id": "c1",
"confidence": 0.8182498514652252,
"choice": "PROMOTE",
"transition_id": "g28e378:446",
"target_state": "g24s145",
"input_confirmed": true,
"_line": 1106
}
},
"selection_at_execution": {
"event": "SELECTION_DECISION",
"step": 450,
"monotonic_ns": 1880283583234694,
"wall_time_ms": 1793991236680,
"abstraction_version": 28,
"state_id": "g28s153",
"model_action_id": "g28a1671",
"snapshot_reason": "selection",
"state_visits": 3,
"state_first_seen": 446,
"state_last_seen": 450,
"state_saturated": false,
"state_saturation": 0.5,
"native_state_saturated": false,
"native_unsaturated_model_action_count": 1,
"model_action_count": 2,
"enabled_valid_action_count": 2,
"unsaturated_model_action_count": 1,
"unvisited_model_action_count": 1,
"effective_unvisited_model_action_count": 1,
"effective_state_saturated": false,
"local_probe_action_count": 0,
"navigation_action_count": 1,
"enabled_valid_unvisited_action_count": 1,
"unvisited_model_action_ids": "g28a1672",
"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": 2,
"selected_action_fingerprint": "type=MODEL_CLICK|activity=org.wikipedia.feed.configure.ConfigureActivity|class=android.widget.TextView|text=Show all|resource_id=org.wikipedia:id/title|bounds=597,139,1028,196",
"selected_resolved_node_count": 3,
"_line": 1102
}
}四、是否值得PROMOTE
无法可靠裁定相对整个Run决策前历史的功能新颖性。
疑似主要环节：UNKNOWN；wire关系审查：DISTINCT_COMPARATORS_VALID。
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