diff --git a/code/SoS/mnm_analysis/mnm_methods/rss_analysis.ipynb b/code/SoS/mnm_analysis/mnm_methods/rss_analysis.ipynb
index 70b578ca0..5ed9fd0b6 100644
--- a/code/SoS/mnm_analysis/mnm_methods/rss_analysis.ipynb
+++ b/code/SoS/mnm_analysis/mnm_methods/rss_analysis.ipynb
@@ -99,7 +99,7 @@
     "\n",
     "- `--methods` (default `susie`), `--coverage` (`0.95`), `--secondary-coverage` (`0.7,0.5`).\n",
     "- `--min-abs-corr` (`0.8`), `--median-abs-corr` (off; sets OR-logic purity), `--pip-cutoff` (`0.025`).\n",
-    "- `--L` (`20`) and `--L-greedy` (`5`): SuSiE fit defaults.\n",
+    "- `--L` (`10`) and `--L-greedy` (`none`; a positive int enables the greedy-L loop): SuSiE fit defaults.\n",
     "- `--method-args`: JSON overriding any of the above per method, e.g. `{\"susie\":{\"L\":10}}`.\n",
     "- `--ser-fallback` / `--no-ser-fallback` (default on), `--r-finite`, `--r-mismatch` (`eb_mix`), `--rss-control`: numerical safeguards for the RSS likelihood.\n",
     "- `--[no-]keep-original-fit` (default off): retain the pre-fallback multi-effect fit on SER-fallback regions (where it differs from the reported SER fit); off keeps none.\n",
@@ -306,8 +306,8 @@
     "  --min-abs-corr 0.8 (as float)\n",
     "  --median-abs-corr ''\n",
     "  --pip-cutoff 0.025 (as float)\n",
-    "  --L 20 (as int)\n",
-    "  --L-greedy 5 (as int)\n",
+    "  --L 10 (as int)\n",
+    "  --L-greedy none (as str; positive int enables greedy-L)\n",
     "  --method-args ''\n",
     "  --[no-]ser-fallback (default to True)\n",
     "                        GWAS SuSiE-RSS fine-mapping: SER fallback is ON by\n",
@@ -392,8 +392,8 @@
     "parameter: min_abs_corr = 0.8\n",
     "parameter: median_abs_corr = ''   # empty -> off (OR-logic purity; needs pecotmr step-1)\n",
     "parameter: pip_cutoff = 0.025\n",
-    "parameter: L = 20\n",
-    "parameter: L_greedy = 5\n",
+    "parameter: L = 10\n",
+    "parameter: L_greedy = 'none'   # 'none'/'off' = greedy off (default); a positive int enables greedy-L\n",
     "parameter: method_args = ''     # nested per-method JSON object\n",
     "# GWAS SuSiE-RSS fine-mapping: SER fallback is ON by default (reproduces the single-panel notebook):\n",
     "# fit finite-sample R + EB LD-mismatch SuSiE-RSS and fall back to the single-effect (SER) result for\n",
diff --git a/code/script/pecotmr_integration/fine_mapping.R b/code/script/pecotmr_integration/fine_mapping.R
index d4a8f5a0b..2b29029a7 100644
--- a/code/script/pecotmr_integration/fine_mapping.R
+++ b/code/script/pecotmr_integration/fine_mapping.R
@@ -95,11 +95,13 @@ parser <- add_argument(parser, "--pip-cutoff",
                        help = "PIP signal cutoff (fineMappingPipeline signalCutoff)",
                        type = "numeric", default = 0.025)
 parser <- add_argument(parser, "--L",
-                       help = "SuSiE number of single effects (susie L); pipeline default 20",
-                       type = "integer", default = 20L)
+                       help = "SuSiE number of single effects (susie L); pipeline default 10",
+                       type = "integer", default = 10L)
 parser <- add_argument(parser, "--L-greedy",
-                       help = "SuSiE greedy init count (susie L_greedy); pipeline default 5",
-                       type = "integer", default = 5L)
+                       help = paste("SuSiE greedy init count (susie L_greedy): a positive integer to run",
+                                    "the greedy-L loop, or 'none'/'off' to disable it; pipeline default",
+                                    "none (greedy off)"),
+                       type = "character", default = "none")
 parser <- add_argument(parser, "--method-args",
                        help = "JSON object {token: {kwarg: value, ...}, ...} for fineMappingPipeline()",
                        type = "character", default = "")
@@ -245,6 +247,22 @@ secondary_cov <- as.numeric(trimws(strsplit(argv$secondary_coverage, ",", fixed
 median_abs_corr <- if (length(argv$median_abs_corr) != 1L || is.na(argv$median_abs_corr))
                      NULL else argv$median_abs_corr
 
+# L_greedy: 'none'/'off'/'null'/empty (case-insensitive) -> NULL, i.e. greedy-L
+# off (susie's own L_greedy default is NULL). Otherwise a positive integer.
+# The NULL is kept via the inline list() in cs_args below, which preserves a
+# named NULL through c()/do.call to fineMappingPipeline; only an x$Lgreedy <- NULL
+# assignment would drop it (re-enabling pecotmr's default 5).
+l_greedy_raw <- trimws(argv[["L_greedy"]])
+l_greedy_val <- if (tolower(l_greedy_raw) %in% c("none", "off", "null", "")) {
+  NULL
+} else {
+  v <- suppressWarnings(as.integer(l_greedy_raw))
+  if (is.na(v) || v < 1L)
+    stop("--L-greedy must be a positive integer or 'none'/'off' (got: '",
+         argv[["L_greedy"]], "')")
+  v
+}
+
 # Seed up front for reproducible fits (mirrors the legacy susie_twas set.seed).
 # NULL when --seed is unset. In QTL mode the value is ALSO forwarded to the
 # pipeline as seed= (below): set.seed() only touches this process's RNG, so it
@@ -302,7 +320,7 @@ methods_arg <- if (is.null(parsed_method_args)) methods else
 # seeds the susie-family tokens and applies explicit --method-args overrides.
 cs_args <- list(methods           = methods_arg,
                 L                 = argv$L,
-                Lgreedy           = argv[["L_greedy"]],
+                Lgreedy           = l_greedy_val,
                 coverage          = argv$coverage,
                 secondaryCoverage = secondary_cov,
                 signalCutoff      = argv$pip_cutoff,