实验 76:A70 10-case Adaptive Router Expansion

保守扩展 adaptive λ router:低冲突/defer case 零误激活,但指标仍是 Chamfer/F@5 tradeoff

实验定位 零误激活 / 指标混合

A70 是一个 10-case retrospective routing audit:不是重新生成 mesh,而是把 A68/A69 在 3 个 active residual case 上的已物化 policy delta 投影回 A51 的 10-case route set;其余 7 个 low-conflict/defer case 强制保持 carrier_control。

关键结果:A70-A68 的 10-case mean ΔChamfer=-0.0001547,优于 A52 的 -0.0000829,但弱于 A53 的 -0.0002088;mean ΔF@5=0.000814,为正但低于 A52。最重要的是 false activation=0,说明 conservative readiness gate 没有污染低冲突样本。

A70 compares A52/A53 selector summaries with A68/A69 replay deltas; active policies were not regenerated in a unified sampler run.

实验设计(Image-2 风格)

实验设计图:A70 10-case router expansion
实验设计图:A70 10-case router expansion
模块设计图:conservative active-only routing
模块设计图:conservative active-only routing

模块设计(Image-2 风格)

输入

  • A51 提供 10-case readiness route:3 个 active residual、7 个 defer。
  • A68 提供 full adaptive router 在 3 个 active case 上的 policy delta。
  • A69 提供 leave-one-out router 的 held-out policy delta。

路由规则

  • active case:读取 A68/A69 policy 输出。
  • defer case:强制 carrier_control,delta=0。
  • 统计 false activation,验证 router 是否污染不该动的 case。
  • 同表比较 A52 original、A53 saturation guard。

实验结果(表格)

方法汇总

selectorcasesactivedefermean ΔChamfermean ΔF@5jointfalse activationpolicy counts
a52_original_selector1037-0.00008290.001326-0.00018900{'carrier_control': 7, 'depth_outward_only': 1, 'depth_signed_strong': 2}
a53_saturation_guard_signed_selector1037-0.00020880.000728-0.00026700{'carrier_control': 8, 'depth_signed_strong': 2}
a70_a68_active_only_router_10case1037-0.00015470.000814-0.00021990{'a66_uncertainty_damped_budget': 1, 'a67_coverage_friendly_damping': 1, 'a67_visible_preserving_damping': 1, 'carrier_control': 7}
a70_a69_loo_active_only_router_10case1037-0.00012990.000717-0.00018730{'a66_uncertainty_damped_budget': 2, 'a67_coverage_friendly_damping': 1, 'carrier_control': 7}

A70-A68 逐 case 路由

caseroute typeA51 routeselected policyΔChamferΔF@5jointsource
gso_000_input4activeready_standard_depth_signa66_uncertainty_damped_budget-0.0003827-0.000773-0.0003208a68_active_policy_pool
gso_001_input4deferready_but_defer_low_conflictcarrier_control0.00000000.0000000.0000000carrier_defer_no_activation
gso_002_input4activeready_high_value_depth_signa67_visible_preserving_damping-0.00108760.007837-0.0017146a68_active_policy_pool
gso_003_input4deferdefer_carrier_not_readycarrier_control0.00000000.0000000.0000000carrier_defer_no_activation
gso_004_input4deferready_but_defer_low_conflictcarrier_control0.00000000.0000000.0000000carrier_defer_no_activation
gso_005_input4deferready_but_defer_low_conflictcarrier_control0.00000000.0000000.0000000carrier_defer_no_activation
gso_006_input4deferdefer_carrier_not_readycarrier_control0.00000000.0000000.0000000carrier_defer_no_activation
gso_007_input4deferready_but_defer_low_conflictcarrier_control0.00000000.0000000.0000000carrier_defer_no_activation
gso_008_input4activeready_high_value_depth_signa67_coverage_friendly_damping-0.00007700.001078-0.0001633a68_active_policy_pool
gso_009_input4deferready_but_defer_low_conflictcarrier_control0.00000000.0000000.0000000carrier_defer_no_activation

可视化结果

A70 Chamfer comparison
A70 Chamfer comparison
A70 F@5 comparison
A70 F@5 comparison
A70 false activation audit
A70 false activation audit

实验结论

下一步想法