Initial commit: Natural Memory Agent Lab:模型服务 + TypeScript agent + 记忆能力场景评测(冲突更新、跨会话回忆、遗忘撤回、多跳、未知拒答)

This commit is contained in:
WpyQwq
2026-09-19 12:09:13 +08:00
commit eb9bd87a4d
108 changed files with 6452 additions and 0 deletions
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
+35
View File
@@ -0,0 +1,35 @@
{
"baseUrl": "http://127.0.0.1:8766/v1",
"outDir": "H:\\Memory\\agent_lab\\agent\\runs\\forget_fixed",
"arms": [
"on",
"off"
],
"scenarios": [
{
"id": "forget_retraction",
"description": "写入事实后明确要求遗忘;跨会话再问时必须不再使用该事实(撤回/覆盖语义),且不得把已忘内容当现状。",
"arms": {
"on": {
"activeRecords": 1,
"assertionsPassed": 0,
"assertionsTotal": 2,
"passRate": 0,
"failures": [
"answer_not_contains: leaked [\"7781\"]",
"answer_contains: missing one of [\"没有\",\"不知道\",\"已\",\"无法确认\",\"未保留\"]"
],
"meanSecondsPerTurn": 30.21
},
"off": {
"activeRecords": 0,
"assertionsPassed": 2,
"assertionsTotal": 2,
"passRate": 1,
"failures": [],
"meanSecondsPerTurn": 38.69
}
}
}
]
}