Updated SBND configurations to fix truth matching issues

Author: francois-drielsma

config/sbnd/README.md

@@ -10,6 +10,31 @@ These weights have been trained using the following files at Polaris:
 
 ...using the configs in the [sbnd_spine_train](https://github.com/bear-is-asleep/sbnd_spine_train/tree/master) repo.
 
+## August 18th 2025
+
+```shell
+sbnd_full_chain_250818.cfg
+sbnd_full_chain_data_250818.cfg
+```
+
+Training samples MPVMPR using `sbndcode v10_04_01` which can be found [here](https://github.com/SBNSoftware/sbndcode/tree/v10_04_01) . The training samples are generated using the following fcls:
+```
+run_mpvmpr_sbnd.fcl
+g4_sce_lite.fcl
+detsim_sce_lite.fcl
+reco1_mpvmpr.fcl
+```
+
+The following modifications were made w.r.t. the March configurations
+- Add a logical containment post-processor
+- Do not move cathode crosser depositions to fix the truth matching
+- Restrict FM interaction candidates to those not clearly out-of-time
+
+Description:
+  - UResNet + PPN + gSPICE + GrapPAs (track + shower + interaction)
+  - Class-weighted loss on PID predictions
+  - The `*_data_*` declination is tailored for data (no labels)
+
 ## March 28th 2025
 
 ```shell

config/sbnd/latest.cfg

@@ -1 +1 @@
-sbnd_full_chain_250328.cfg
+sbnd_full_chain_250818.cfg

config/sbnd/latest_data.cfg

@@ -1 +1 @@
-sbnd_full_chain_data_250328.cfg
+sbnd_full_chain_data_250818.cfg

config/sbnd/sbnd_full_chain_250328.cfg

@@ -467,7 +467,7 @@ build:
   fragments: false
   particles: true
   interactions: true
-  
+
 # Run post-processors
 post:
   cathode_crosser:
@@ -476,8 +476,8 @@ post:
     crossing_point_tolerance: 565.685
     offset_tolerance: 15
     angle_tolerance: 0.26
-    adjust_crossers: True
-    merge_crossers: True
+    adjust_crossers: true
+    merge_crossers: true
     priority: 4
   flash_match:
     flash_key: flashes