Confirm v3 rejection with extended fixed-MATH alpha line search

candidates/budgie-alignment-v2/reasoning-sft-then-rl-v3/recovery/V3_FINAL_DECISION.json

@@ -94,8 +94,19 @@
       }
     }
   },
-  "decision": "Reject reasoning SFT Stage A because the only alpha with an information-management gain regresses fixed GSM and MATH. Do not run follow-on RL. Smaller alphas have no held-out information gain. Keep verified-math-a025.",
+  "decision": "Reject Stage A. An extended task-vector search from alpha .005 through 1.0 found no candidate matching the leader fixed MATH score 3/15; best candidates reach 2/15. Stage B RL is not run.",
   "parameter_count_change": 0,
   "benchmark_family_rows": 0,
-  "public_root_changed": false
+  "public_root_changed": false,
+  "fixed_math_task_vector_line_search": {
+    "leader": 3,
+    "alpha_0_005": 2,
+    "alpha_0_025": 1,
+    "alpha_0_05": 1,
+    "alpha_0_1": 2,
+    "alpha_0_2": 2,
+    "alpha_0_5": 2,
+    "alpha_1_0": 2,
+    "result": "No tested alpha matches the leader."
+  }
 }