diff --git a/src/models/gemma3.cpp b/src/models/gemma3.cpp
index ae60ef4790c9..1ad472b449c1 100644
--- a/src/models/gemma3.cpp
+++ b/src/models/gemma3.cpp
@@ -33,7 +33,7 @@ llm_build_gemma3<iswa>::llm_build_gemma3(const llama_model & model, const llm_gr
         float freq_base_l  = 0.0f;
         float freq_scale_l = 0.0f;
 
-        if constexpr (iswa) {
+        if (hparams.is_swa(il)) {
             freq_base_l  = model.get_rope_freq_base (cparams, il);
             freq_scale_l = model.get_rope_freq_scale(cparams, il);
         } else {
diff --git a/src/models/modern-bert.cpp b/src/models/modern-bert.cpp
index c7809bdedfa8..559b119ed7fa 100644
--- a/src/models/modern-bert.cpp
+++ b/src/models/modern-bert.cpp
@@ -26,6 +26,7 @@ llm_build_modern_bert<iswa>::llm_build_modern_bert(const llama_model & model, co
     for (int il = 0; il < n_layer; ++il) {
         float freq_base_l  = 0.0f;
 
+        // TODO @ngxson : do we expect to apply this globally, or dynamically per layer?
         if constexpr (iswa) {
             freq_base_l = model.get_rope_freq_base(cparams, il);
         } else {
diff --git a/src/models/plamo3.cpp b/src/models/plamo3.cpp
index 55c8064679e2..7a9468cd33e0 100644
--- a/src/models/plamo3.cpp
+++ b/src/models/plamo3.cpp
@@ -26,7 +26,7 @@ llm_build_plamo3<iswa>::llm_build_plamo3(const llama_model & model, const llm_gr
 
         float freq_base_l  = 0.0f;
         float freq_scale_l = 0.0f;
-        if constexpr (iswa) {
+        if (hparams.is_swa(il)) {
             freq_base_l  = model.get_rope_freq_base (cparams, il);
             freq_scale_l = model.get_rope_freq_scale(cparams, il);
         } else {