diag: H 段（老师 oracle 走 rollout / 前缀消融 / 子目标越过顶点规则 / 分轴与分类别指标）；训练：EMA、选模不落在 p_neq 预热、闭环分数选模、24000 步余弦、p_neq 0.5 + p_edge 0.1、子目标头切断抄输入、前缀/dq 增广；仿真：last 记真正提交的候选、碰撞与到达按每题时限、回放门禁读 pass+指纹并走 rollout；数据：动作标签按 1.25·STEP 归一化（act_scale 随 meta/ckpt/ONNX 传递，旧模型 1.0）、收回亏空结转、DAgger 带 qPrev 与起步速度、打包签名变更自动重打；ik-policy.js 同步（actScale、子目标规则、DLS 时清 lastPlan）；本机执行清单与 README 更新

ml/configs/base.json

@@ -1,10 +1,11 @@
 {
   "model": {"d": 192, "heads": 8, "ff": 768, "enc": 4, "dec": 4},
   "train": {
-    "batch": 1024, "lr": 1e-3, "warmup": 1000, "steps": 60000, "wd": 0.01, "betas": [0.9, 0.95],
+    "batch": 1024, "lr": 1e-3, "warmup": 1000, "steps": 24000, "wd": 0.01, "betas": [0.9, 0.95],
     "amp": "bf16", "grad_clip": 1.0,
     "w_mf": 1.0, "w_fk": 0.5, "w_smooth": 0.05, "w_value": 0.1, "w_sub": 0.1,
-    "p_neq": 0.25, "p_neq_warm": 5000, "fk_tmax": 0.3, "fk_caps": 96,
+    "p_neq": 0.5, "p_edge": 0.1, "p_neq_warm": 5000, "fk_tmax": 0.3, "fk_caps": 96,
+    "ema": 0.999, "aug_pre_drop": 0.15, "aug_pre_sigma": 0.1, "aug_dq_sigma": 0.3,
     "ckpt_min": 10, "eval_every": 2000, "eval_tasks": 512, "log_every": 50, "patience_evals": 5
   }
 }