diff --git a/bin/fink b/bin/fink
index 85d01580..d24d95af 100755
--- a/bin/fink
+++ b/bin/fink
@@ -505,6 +505,7 @@ elif [[ $service == "hostless" ]]; then
   ${FINK_HOME}/bin/hostless_detection.py ${HELP_ON_SERVICE} \
   -agg_data_prefix ${AGG_DATA_PREFIX} \
   -night ${NIGHT} \
+  -hostless_folder "hostless_ids" \
   -log_level ${LOG_LEVEL} ${EXIT_AFTER}
 elif [[ $service == "dwarf_agn" ]]; then
   spark-submit --master ${SPARK_MASTER} \
diff --git a/bin/generate_ssoft.py b/bin/generate_ssoft.py
index a7e3de3e..182dc347 100644
--- a/bin/generate_ssoft.py
+++ b/bin/generate_ssoft.py
@@ -21,7 +21,7 @@
 from fink_broker.logging_utils import get_fink_logger, inspect_application
 from fink_broker.spark_utils import init_sparksession
 
-from fink_spins.ssoft import build_the_ssoft
+from fink_science.ssoft.processor import build_the_ssoft
 
 
 def main():
@@ -81,7 +81,7 @@ def main():
 
     # Initialise Spark session
     spark = init_sparksession(
-        name="ssoft_{}_{}".format(args.model, version), shuffle_partitions=20
+        name="ssoft_{}_{}".format(args.model, version), shuffle_partitions=200
     )
 
     # The level here should be controlled by an argument.
@@ -106,6 +106,7 @@ def main():
         frac=args.frac,
         model=args.model,
         version=version,
+        sb_method="fastnifty",
     )
 
     pdf.to_parquet("ssoft_{}_{}.parquet".format(args.model, version))
diff --git a/bin/hostless_detection.py b/bin/hostless_detection.py
index c3a29b46..616700a0 100644
--- a/bin/hostless_detection.py
+++ b/bin/hostless_detection.py
@@ -17,7 +17,7 @@
 
 import argparse
 import os
-
+import pandas as pd
 
 from pyspark.sql import functions as F
 
@@ -33,6 +33,8 @@
 from fink_utils.tg_bot.utils import get_cutout
 from fink_utils.tg_bot.utils import msg_handler_tg_cutouts
 
+from fink_tns.utils import read_past_ids
+
 from fink_science.hostless_detection.processor import run_potential_hostless
 
 
@@ -119,10 +121,20 @@ def main():
     cond_template = df["kstest_static"][1] <= 0.85
     pdf = df.filter(cond_science).filter(cond_template).select(cols_).toPandas()
 
+    # load hostless IDs
+    past_ids = read_past_ids(args.hostless_folder)
+
+    new_ids = []
     # Loop over matches & send to Telegram
     if ("FINK_TG_TOKEN" in os.environ) and os.environ["FINK_TG_TOKEN"] != "":
         payloads = []
         for _, alert in pdf.iterrows():
+            # Do not send request if the object
+            # has been already reported by the bot
+            if alert["objectId"] in past_ids.to_numpy():
+                print("{} already sent!".format(alert["objectId"]))
+                continue
+
             curve_png = get_curve(
                 objectId=alert["objectId"],
                 origin="API",
@@ -144,10 +156,17 @@ def main():
             )
 
             payloads.append((text, curve_png, [cutout_science, cutout_template]))
+            new_ids.append(alert["objectId"])
 
         if len(payloads) > 0:
+            # Send to tg
             msg_handler_tg_cutouts(payloads, channel_id="@fink_hostless", init_msg="")
 
+            # Save ids on disk
+            pdf_ids = pd.DataFrame.from_dict({"id": new_ids})
+            name = "{}{}{}".format(args.night[:4], args.night[4:6], args.night[6:8])
+            pdf_ids.to_csv("{}/{}.csv".format(args.hostless_folder, name), index=False)
+
 
 if __name__ == "__main__":
     main()
diff --git a/bin/raw2science_batch.py b/bin/raw2science_batch.py
index 33423c02..66347250 100644
--- a/bin/raw2science_batch.py
+++ b/bin/raw2science_batch.py
@@ -67,7 +67,7 @@ def main():
     df = df.filter(df["candidate.nbad"] == 0).filter(df["candidate.rb"] >= 0.55)
 
     # Apply science modules
-    df = apply_science_modules(df, logger)
+    df = apply_science_modules(df)
 
     # Add tracklet information
     df_trck = spark.read.format("parquet").load(input_raw)
diff --git a/deps/requirements-science.txt b/deps/requirements-science.txt
index d499911c..e66a28e6 100644
--- a/deps/requirements-science.txt
+++ b/deps/requirements-science.txt
@@ -1,13 +1,13 @@
 # Fink-science dependencies
 # WARNING: fink-science pip module is deprecated and fink-science is now installed from source
 
-git+https://github.com/astrolabsoftware/fink-science@5.19.0
+git+https://github.com/astrolabsoftware/fink-science@5.20.0
 
 # xmatch_cds
 line_profiler==4.1.3
 
 # Active learning
--e git+https://github.com/emilleishida/fink_sn_activelearning.git@4f46b3a1e29de45793125452974e71e92c1ea454#egg=actsnfink
+-e git+https://github.com/emilleishida/fink_sn_activelearning.git@cb61bbb34630c7811862050389b0f993de9639c7#egg=actsnfink
 -e git+https://github.com/COINtoolbox/ActSNClass.git@2c61da91a9d13834d39804fc35aeb3245ba20755#egg=actsnclass
 joblib==1.2.0
 
@@ -31,8 +31,7 @@ iminuit==2.21.0
 -e git+https://github.com/b-biswas/kndetect@kndetect#egg=kndetect
 
 # CATS dependencies
-tensorflow==2.9.2
-tensorflow_addons==0.18.0
+tensorflow==2.15
 
 # Anomalies
 onnxruntime
diff --git a/deps/requirements.txt b/deps/requirements.txt
index 6d90d85c..0da71837 100644
--- a/deps/requirements.txt
+++ b/deps/requirements.txt
@@ -19,7 +19,7 @@ fastavro==1.6.0
 
 # Fink core
 fink_filters>=3.37
-fink-utils>=0.25.0
+fink-utils>=0.28.0
 fink-spins>=0.3.8
 fink-tns>=0.9
 
diff --git a/fink_broker/parser.py b/fink_broker/parser.py
index 0f51a319..bf94a7b7 100644
--- a/fink_broker/parser.py
+++ b/fink_broker/parser.py
@@ -350,6 +350,15 @@ def getargs(parser: argparse.ArgumentParser) -> argparse.Namespace:
         [TNS_FOLDER]
         """,
     )
+    parser.add_argument(
+        "-hostless_folder",
+        type=str,
+        default="hostless_ids",
+        help="""
+        Folder to store ids for hostless detection
+        [HOSTLESS_FOLDER]
+        """,
+    )
     parser.add_argument(
         "--tns_sandbox",
         action="store_true",
diff --git a/fink_broker/science.py b/fink_broker/science.py
index d97f9960..8e395e1c 100644
--- a/fink_broker/science.py
+++ b/fink_broker/science.py
@@ -46,8 +46,7 @@
 from fink_science.random_forest_snia.processor import rfscore_rainbow_elasticc
 from fink_science.snn.processor import snn_ia_elasticc, snn_broad_elasticc
 from fink_science.cats.processor import predict_nn
-from fink_science.agn.processor import agn_elasticc
-from fink_science.slsn.processor import slsn_elasticc
+from fink_science.slsn.processor import slsn_elasticc_with_md
 from fink_science.fast_transient_rate.processor import magnitude_rate
 from fink_science.fast_transient_rate import rate_module_output_schema
 # from fink_science.t2.processor import t2
@@ -599,7 +598,7 @@ def apply_science_modules_elasticc(df: DataFrame) -> DataFrame:
     df = df.withColumn("cats_broad_class", mapping_cats_general_expr[df["cats_argmax"]])
     df = df.withColumnRenamed("cbpf_preds", "cats_broad_array_prob")
 
-    # AGN & SLSN
+    # SLSN
     args_forced = [
         "diaObject.diaObjectId",
         "cmidPointTai",
@@ -610,11 +609,9 @@ def apply_science_modules_elasticc(df: DataFrame) -> DataFrame:
         "diaSource.decl",
         "diaObject.hostgal_zphot",
         "diaObject.hostgal_zphot_err",
-        "diaObject.hostgal_ra",
-        "diaObject.hostgal_dec",
+        "diaObject.hostgal_snsep",
     ]
-    df = df.withColumn("rf_agn_vs_nonagn", agn_elasticc(*args_forced))
-    df = df.withColumn("rf_slsn_vs_nonslsn", slsn_elasticc(*args_forced))
+    df = df.withColumn("rf_slsn_vs_nonslsn", slsn_elasticc_with_md(*args_forced))
 
     # Drop temp columns
     df = df.drop(*expanded)