Auto-sync: 2026-06-30 23:06:36

results/paper_analysis.json

@@ -1,6 +1,6 @@
 {
   "best_clean_key": "residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmchallenger001",
-  "generated_utc": "2026-07-01T01:39:11+00:00",
+  "generated_utc": "2026-07-01T03:07:08+00:00",
   "mechanism_gap": {
     "best_clean_vs_direct_same_ckpt": 0.0776811594202898,
     "best_clean_vs_h16": 0.06318840579710144,
@@ -2208,6 +2208,11 @@
       "source": "results/h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmchallenger0p03_summary.json",
       "std_success": 0.012819330079707819
     },
+    "residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_noopwgcontact_challenger001": {
+      "label": "K4 compatible tangents, no-op bonus 0.03, near-miss global + no-op/wrong-gripper contact challenger",
+      "missing": true,
+      "source": "results/h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_noopwgcontact_challenger0p01_summary.json"
+    },
     "residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_challenger001": {
       "ci95_success": 0.03920403094295293,
       "label": "K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper challenger on Pick/Pull, gate 0.01",
@@ -2306,14 +2311,203 @@
       "std_success": 0.015780472566743406
     },
     "residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin003_challenger001": {
+      "ci95_success": 0.039598844840038006,
       "label": "K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper Pick/Pull margin 0.03",
-      "missing": true,
-      "source": "results/h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpull_wgmargin0p03_challenger0p01_summary.json"
+      "mean_action_mse_to_best": 0.41668552688795374,
+      "mean_progress": 0.5738671411409219,
+      "mean_success": 0.36000000000000004,
+      "num_completed": 3,
+      "per_task_success": {
+        "LiftPegUpright-v1": {
+          "mean_num_groups": 102.0,
+          "mean_success": 0.26340865087329823,
+          "std_success": 0.06057881385012102
+        },
+        "PickCube-v1": {
+          "mean_num_groups": 196.66666666666666,
+          "mean_success": 0.3289542583020844,
+          "std_success": 0.04030135158178743
+        },
+        "PullCube-v1": {
+          "mean_num_groups": 81.0,
+          "mean_success": 0.228271436166173,
+          "std_success": 0.03525140677829959
+        },
+        "PushCube-v1": {
+          "mean_num_groups": 102.0,
+          "mean_success": 0.7604066042026859,
+          "std_success": 0.08151850231627764
+        },
+        "StackCube-v1": {
+          "mean_num_groups": 93.33333333333333,
+          "mean_success": 0.19707329707329704,
+          "std_success": 0.0831467196651622
+        }
+      },
+      "seed_action_mse_to_best": {
+        "0": 0.39585299389679796,
+        "1": 0.40845387818858675,
+        "2": 0.44574970857847646
+      },
+      "seed_progress": {
+        "0": 0.5592340455844027,
+        "1": 0.5719369965759309,
+        "2": 0.590430381262432
+      },
+      "seed_success": {
+        "0": 0.3565217391304348,
+        "1": 0.34608695652173915,
+        "2": 0.3773913043478261
+      },
+      "selected_candidate_type_counts": {
+        "retrieval_residual_policy_residual": 1476,
+        "retrieval_residual_residual_near_miss": 91,
+        "retrieval_residual_residual_near_miss+residual_wrong_gripper": 5,
+        "retrieval_residual_residual_no_op": 38,
+        "retrieval_residual_residual_no_op+residual_wrong_gripper": 27,
+        "retrieval_residual_residual_wrong_gripper": 88
+      },
+      "selected_residual_scale_counts": {
+        "0.35": 1530,
+        "0.4": 29,
+        "0.45": 166
+      },
+      "selected_type_outcomes": {
+        "retrieval_residual_policy_residual": {
+          "count": 1476.0,
+          "mean_progress": 0.5566651973590068,
+          "success_rate": 0.3346883468834688
+        },
+        "retrieval_residual_residual_near_miss": {
+          "count": 91.0,
+          "mean_progress": 0.5661385127926594,
+          "success_rate": 0.38461538461538464
+        },
+        "retrieval_residual_residual_near_miss+residual_wrong_gripper": {
+          "count": 5.0,
+          "mean_progress": 0.7557853102684021,
+          "success_rate": 0.6
+        },
+        "retrieval_residual_residual_no_op": {
+          "count": 38.0,
+          "mean_progress": 0.7757954887653652,
+          "success_rate": 0.5263157894736842
+        },
+        "retrieval_residual_residual_no_op+residual_wrong_gripper": {
+          "count": 27.0,
+          "mean_progress": 0.6916872307106301,
+          "success_rate": 0.48148148148148145
+        },
+        "retrieval_residual_residual_wrong_gripper": {
+          "count": 88.0,
+          "mean_progress": 0.7367008198687638,
+          "success_rate": 0.6363636363636364
+        }
+      },
+      "source": "results/h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpull_wgmargin0p03_challenger0p01_summary.json",
+      "std_success": 0.01593939372158552
     },
     "residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin005_challenger001": {
+      "ci95_success": 0.03920403094295293,
       "label": "K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper Pick/Pull margin 0.05",
+      "mean_action_mse_to_best": 0.41403828773242624,
+      "mean_progress": 0.5730632333303599,
+      "mean_success": 0.3588405797101449,
+      "num_completed": 3,
+      "per_task_success": {
+        "LiftPegUpright-v1": {
+          "mean_num_groups": 102.0,
+          "mean_success": 0.26340865087329823,
+          "std_success": 0.06057881385012102
+        },
+        "PickCube-v1": {
+          "mean_num_groups": 196.66666666666666,
+          "mean_success": 0.3272707566185827,
+          "std_success": 0.03916710518683381
+        },
+        "PullCube-v1": {
+          "mean_num_groups": 81.0,
+          "mean_success": 0.22388547125389233,
+          "std_success": 0.04131573991107309
+        },
+        "PushCube-v1": {
+          "mean_num_groups": 102.0,
+          "mean_success": 0.7604066042026859,
+          "std_success": 0.08151850231627764
+        },
+        "StackCube-v1": {
+          "mean_num_groups": 93.33333333333333,
+          "mean_success": 0.19707329707329704,
+          "std_success": 0.0831467196651622
+        }
+      },
+      "seed_action_mse_to_best": {
+        "0": 0.3946219221810284,
+        "1": 0.4068148876516067,
+        "2": 0.4406780533646436
+      },
+      "seed_progress": {
+        "0": 0.5592912306045384,
+        "1": 0.5702725335210562,
+        "2": 0.5896259358654852
+      },
+      "seed_success": {
+        "0": 0.3565217391304348,
+        "1": 0.3443478260869565,
+        "2": 0.37565217391304345
+      },
+      "selected_candidate_type_counts": {
+        "retrieval_residual_policy_residual": 1503,
+        "retrieval_residual_residual_near_miss": 91,
+        "retrieval_residual_residual_near_miss+residual_wrong_gripper": 5,
+        "retrieval_residual_residual_no_op": 38,
+        "retrieval_residual_residual_no_op+residual_wrong_gripper": 27,
+        "retrieval_residual_residual_wrong_gripper": 61
+      },
+      "selected_residual_scale_counts": {
+        "0.35": 1550,
+        "0.4": 26,
+        "0.45": 149
+      },
+      "selected_type_outcomes": {
+        "retrieval_residual_policy_residual": {
+          "count": 1503.0,
+          "mean_progress": 0.5571624199836251,
+          "success_rate": 0.3366600133067199
+        },
+        "retrieval_residual_residual_near_miss": {
+          "count": 91.0,
+          "mean_progress": 0.5661385127926594,
+          "success_rate": 0.38461538461538464
+        },
+        "retrieval_residual_residual_near_miss+residual_wrong_gripper": {
+          "count": 5.0,
+          "mean_progress": 0.7557853102684021,
+          "success_rate": 0.6
+        },
+        "retrieval_residual_residual_no_op": {
+          "count": 38.0,
+          "mean_progress": 0.7757954887653652,
+          "success_rate": 0.5263157894736842
+        },
+        "retrieval_residual_residual_no_op+residual_wrong_gripper": {
+          "count": 27.0,
+          "mean_progress": 0.6916872307106301,
+          "success_rate": 0.48148148148148145
+        },
+        "retrieval_residual_residual_wrong_gripper": {
+          "count": 61.0,
+          "mean_progress": 0.7814040203563503,
+          "success_rate": 0.6885245901639344
+        }
+      },
+      "source": "results/h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpull_wgmargin0p05_challenger0p01_summary.json",
+      "std_success": 0.015780472566743406
+    },
+    "residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpullstack_challenger001": {
+      "label": "K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper challenger on Pick/Pull/Stack",
       "missing": true,
-      "source": "results/h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpull_wgmargin0p05_challenger0p01_summary.json"
+      "source": "results/h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpullstack_challenger0p01_summary.json"
     },
     "residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmwgchallenger001": {
       "ci95_success": 0.028111505659259583,

results/paper_analysis.md

@@ -1,6 +1,6 @@
 # Paper Analysis
 
-Generated: `2026-07-01T01:39:11+00:00`
+Generated: `2026-07-01T03:07:08+00:00`
 
 ## Main Seed Statistics
 
@@ -65,8 +65,10 @@ Generated: `2026-07-01T01:39:11+00:00`
 | residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmwgmargin005_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss challenger 0.01, wrong-gripper margin 0.05 | 3 | 35.59% +/- 1.31 | +/- 3.24 | 57.14% | 0.417 | +5.86 pp |
 | residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmwgchallenger001_pickpull | K4 compatible tangents, no-op bonus 0.03, near-miss/wrong-gripper challenger gate 0.01 on Pick/Pull only | 3 | 36.00% +/- 1.66 | +/- 4.12 | 57.36% | 0.418 | +6.26 pp |
 | residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper challenger on Pick/Pull, gate 0.01 | 3 | 36.06% +/- 1.58 | +/- 3.92 | 57.41% | 0.419 | +6.32 pp |
-| residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin003_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper Pick/Pull margin 0.03 | 0 | missing | missing | missing | missing | missing |
-| residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin005_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper Pick/Pull margin 0.05 | 0 | missing | missing | missing | missing | missing |
+| residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin003_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper Pick/Pull margin 0.03 | 3 | 36.00% +/- 1.59 | +/- 3.96 | 57.39% | 0.417 | +6.26 pp |
+| residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin005_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper Pick/Pull margin 0.05 | 3 | 35.88% +/- 1.58 | +/- 3.92 | 57.31% | 0.414 | +6.14 pp |
+| residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpullstack_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper challenger on Pick/Pull/Stack | 0 | missing | missing | missing | missing | missing |
+| residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_noopwgcontact_challenger001 | K4 compatible tangents, no-op bonus 0.03, near-miss global + no-op/wrong-gripper contact challenger | 0 | missing | missing | missing | missing | missing |
 | typed_proposal_lattice_types6_prepend_margin000 | Typed proposal lattice head, six families, policy-prepended margin 0.00 | 3 | 31.30% +/- 3.81 | +/- 9.48 | 54.03% | 0.480 | +1.57 pp |
 | typed_proposal_lattice_types6_prepend_margin005 | Typed proposal lattice head, six families, policy-prepended margin 0.05 | 3 | 32.17% +/- 3.05 | +/- 7.57 | 54.74% | 0.467 | +2.43 pp |
 | typed_proposal_lattice_types4safe_prepend_margin005 | Typed proposal lattice head, safe four families, policy-prepended margin 0.05 | 3 | 33.45% +/- 3.39 | +/- 8.43 | 55.47% | 0.456 | +3.71 pp |

results/paper_table_status.json

@@ -1648,6 +1648,44 @@
       "story_role": "type-specific affordance margin: keep near-miss globally sensitive while requiring stronger wrong-gripper evidence only on Pick/Pull",
       "fallback_success": null,
       "pending_job": "14991117/14991125",
+      "path_exists": true,
+      "status": "complete",
+      "success": 0.36,
+      "std_success": 0.01593939372158552,
+      "completed_seeds": null,
+      "num_completed": 3,
+      "best_config": null,
+      "gain_vs_h16_policy": 0.06260869565217392
+    },
+    {
+      "key": "retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin005_challenger001",
+      "label": "K4 compatible residual retrieval, near-miss global + wrong-gripper Pick/Pull margin 0.05",
+      "path": "h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpull_wgmargin0p05_challenger0p01_summary.json",
+      "clean_deployment": "yes",
+      "same_state_proposals": "no",
+      "expert_proposal": "no",
+      "story_role": "upper setting for type-specific wrong-gripper evidence on Pick/Pull while preserving global near-miss calibration",
+      "fallback_success": null,
+      "pending_job": "14991123/14991126",
+      "path_exists": true,
+      "status": "complete",
+      "success": 0.35884057971014494,
+      "std_success": 0.015780472566743406,
+      "completed_seeds": null,
+      "num_completed": 3,
+      "best_config": null,
+      "gain_vs_h16_policy": 0.06144927536231887
+    },
+    {
+      "key": "retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpullstack_challenger001",
+      "label": "K4 compatible residual retrieval, near-miss global + wrong-gripper challenger on Pick/Pull/Stack",
+      "path": "h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpullstack_challenger0p01_summary.json",
+      "clean_deployment": "yes",
+      "same_state_proposals": "no",
+      "expert_proposal": "no",
+      "story_role": "oracle-trace contact-task calibration: wrong-gripper residuals are only challenger-eligible on Pick/Pull/Stack while near-miss stays global",
+      "fallback_success": null,
+      "pending_job": "14995028/14995047",
       "path_exists": false,
       "status": "pending",
       "success": null,
@@ -1658,15 +1696,15 @@
       "gain_vs_h16_policy": null
     },
     {
-      "key": "retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin005_challenger001",
-      "label": "K4 compatible residual retrieval, near-miss global + wrong-gripper Pick/Pull margin 0.05",
-      "path": "h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpull_wgmargin0p05_challenger0p01_summary.json",
+      "key": "retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_noopwgcontact_challenger001",
+      "label": "K4 compatible residual retrieval, near-miss global + no-op/wrong-gripper contact challenger",
+      "path": "h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_noopwgcontact_challenger0p01_summary.json",
       "clean_deployment": "yes",
       "same_state_proposals": "no",
       "expert_proposal": "no",
-      "story_role": "upper setting for type-specific wrong-gripper evidence on Pick/Pull while preserving global near-miss calibration",
+      "story_role": "contact-task sparse-family challenger ablation using no-op and wrong-gripper residuals where oracle trace shows proposal headroom",
       "fallback_success": null,
-      "pending_job": "14991123/14991126",
+      "pending_job": "14995045/14995048",
       "path_exists": false,
       "status": "pending",
       "success": null,

results/paper_table_status.md

@@ -89,8 +89,10 @@ Baseline h=16 policy: 29.74%
 | retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmwgmargin005_challenger001 | K4 compatible residual retrieval, near-miss challenger 0.01, wrong-gripper margin 0.05 | complete | 35.59% | +5.86 pp | yes | no | no | family-specific reliability margin upper setting for wrong-gripper challenger evidence |
 | retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmwgchallenger001_pickpull | K4 compatible residual retrieval, near-miss/wrong-gripper challenger gate 0.01 on Pick/Pull only | complete | 36.00% | +6.26 pp | yes | no | no | task-conditional challenger gate using branch trace evidence that wrong-gripper tangents help Pick/Pull but harm other tasks |
 | retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_challenger001 | K4 compatible residual retrieval, near-miss global + wrong-gripper challenger on Pick/Pull, gate 0.01 | complete | 36.06% | +6.32 pp | yes | no | no | type-specific affordance gate: keep near-miss global while restricting wrong-gripper challenger to acquisition/contact tasks |
-| retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin003_challenger001 | K4 compatible residual retrieval, near-miss global + wrong-gripper Pick/Pull margin 0.03 | pending 14991117/14991125 | pending | pending | yes | no | no | type-specific affordance margin: keep near-miss globally sensitive while requiring stronger wrong-gripper evidence only on Pick/Pull |
-| retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin005_challenger001 | K4 compatible residual retrieval, near-miss global + wrong-gripper Pick/Pull margin 0.05 | pending 14991123/14991126 | pending | pending | yes | no | no | upper setting for type-specific wrong-gripper evidence on Pick/Pull while preserving global near-miss calibration |
+| retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin003_challenger001 | K4 compatible residual retrieval, near-miss global + wrong-gripper Pick/Pull margin 0.03 | complete | 36.00% | +6.26 pp | yes | no | no | type-specific affordance margin: keep near-miss globally sensitive while requiring stronger wrong-gripper evidence only on Pick/Pull |
+| retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpull_wgmargin005_challenger001 | K4 compatible residual retrieval, near-miss global + wrong-gripper Pick/Pull margin 0.05 | complete | 35.88% | +6.14 pp | yes | no | no | upper setting for type-specific wrong-gripper evidence on Pick/Pull while preserving global near-miss calibration |
+| retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpullstack_challenger001 | K4 compatible residual retrieval, near-miss global + wrong-gripper challenger on Pick/Pull/Stack | pending 14995028/14995047 | pending | pending | yes | no | no | oracle-trace contact-task calibration: wrong-gripper residuals are only challenger-eligible on Pick/Pull/Stack while near-miss stays global |
+| retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_noopwgcontact_challenger001 | K4 compatible residual retrieval, near-miss global + no-op/wrong-gripper contact challenger | pending 14995045/14995048 | pending | pending | yes | no | no | contact-task sparse-family challenger ablation using no-op and wrong-gripper residuals where oracle trace shows proposal headroom |
 | typed_proposal_lattice_types6_prepend_margin000 | Typed proposal lattice head, six families, policy-prepended margin 0.00 | complete | 31.30% | +1.57 pp | yes | no | no | model-generated counterfactual proposal support test; field scores typed proposals rather than retrieved train-state residuals |
 | typed_proposal_lattice_types6_prepend_margin005 | Typed proposal lattice head, six families, policy-prepended margin 0.05 | complete | 32.17% | +2.43 pp | yes | no | no | model-generated counterfactual proposal support test with a conservative policy-abstention margin |
 | typed_proposal_lattice_types4safe_prepend_margin005 | Typed proposal lattice head, safe four families, policy-prepended margin 0.05 | complete | 33.45% | +3.71 pp | yes | no | no | anti-goal proposal-family mask test for the model-generated typed lattice |

scripts/build_paper_analysis.py

@@ -513,6 +513,22 @@ METHODS = [
             "k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpull_wgmargin0p05_challenger0p01_summary.json"
         ),
     ),
+    MethodSpec(
+        key="residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpullstack_challenger001",
+        label="K4 compatible tangents, no-op bonus 0.03, near-miss global + wrong-gripper challenger on Pick/Pull/Stack",
+        summary_path=(
+            "h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_"
+            "k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpullstack_challenger0p01_summary.json"
+        ),
+    ),
+    MethodSpec(
+        key="residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_noopwgcontact_challenger001",
+        label="K4 compatible tangents, no-op bonus 0.03, near-miss global + no-op/wrong-gripper contact challenger",
+        summary_path=(
+            "h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_"
+            "k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_noopwgcontact_challenger0p01_summary.json"
+        ),
+    ),
     MethodSpec(
         key="typed_proposal_lattice_types6_prepend_margin000",
         label="Typed proposal lattice head, six families, policy-prepended margin 0.00",

scripts/build_paper_table_status.py

@@ -895,6 +895,26 @@ SPECS = [
         story_role="upper setting for type-specific wrong-gripper evidence on Pick/Pull while preserving global near-miss calibration",
         pending_job="14991123/14991126",
     ),
+    ResultSpec(
+        key="retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_wgpickpullstack_challenger001",
+        label="K4 compatible residual retrieval, near-miss global + wrong-gripper challenger on Pick/Pull/Stack",
+        path="h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_wgpickpullstack_challenger0p01_summary.json",
+        clean_deployment="yes",
+        same_state_proposals="no",
+        expert_proposal="no",
+        story_role="oracle-trace contact-task calibration: wrong-gripper residuals are only challenger-eligible on Pick/Pull/Stack while near-miss stays global",
+        pending_job="14995028/14995047",
+    ),
+    ResultSpec(
+        key="retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_noopbonus003_nmglobal_noopwgcontact_challenger001",
+        label="K4 compatible residual retrieval, near-miss global + no-op/wrong-gripper contact challenger",
+        path="h16_policy_ckpt_near_miss_policy_bc5_bestpt_retrieval_residual_k4_composemasked_dropnmnoop_grid035040045_safe_margin0p20_noopbonus0p03_nmglobal_noopwgcontact_challenger0p01_summary.json",
+        clean_deployment="yes",
+        same_state_proposals="no",
+        expert_proposal="no",
+        story_role="contact-task sparse-family challenger ablation using no-op and wrong-gripper residuals where oracle trace shows proposal headroom",
+        pending_job="14995045/14995048",
+    ),
     ResultSpec(
         key="typed_proposal_lattice_types6_prepend_margin000",
         label="Typed proposal lattice head, six families, policy-prepended margin 0.00",