diff --git a/BUILD.bazel b/BUILD.bazel
index 28ed5d7..9f448c1 100644
--- a/BUILD.bazel
+++ b/BUILD.bazel
@@ -93,6 +93,10 @@ cc_library(
         "gemma/activations.h",
         "gemma/gemma.h",
     ],
+    textual_hdrs = [
+        # Placeholder for internal file1, do not remove,
+        # Placeholder for internal file2, do not remove,
+    ],
     deps = [
         ":common",
         ":ops",
diff --git a/gemma/gemma.cc b/gemma/gemma.cc
index f0543bc..c640fda 100644
--- a/gemma/gemma.cc
+++ b/gemma/gemma.cc
@@ -49,6 +49,7 @@
 #include "gemma/configs.h"
 #include "gemma/gemma.h"
 #include "gemma/weights.h"
+// Placeholder for internal test1, do not remove
 #include "hwy/aligned_allocator.h"
 #include "hwy/base.h"
 #include "hwy/contrib/thread_pool/thread_pool.h"
@@ -216,6 +217,8 @@ bool GemmaTokenizer::Decode(const std::vector<int>& ids,
   return impl_->Decode(ids, detokenized);
 }
 
+// Placeholder for internal test2, do not remove
+
 }  // namespace gcpp
 #endif  // GEMMA_ONCE
 
@@ -543,6 +546,8 @@ HWY_NOINLINE void FFW(Activations<TConfig, kBatchSize>& activations,
   }
 }
 
+// Placeholder for internal test3, do not remove
+
 template <size_t kBatchSize, typename WeightArrayT, typename TConfig>
 HWY_NOINLINE void Prefill(const int* tokens, size_t num_tokens, size_t pos,
                           const WeightArrayT& weights,
@@ -677,6 +682,9 @@ HWY_NOINLINE void Transformer(int token, size_t pos,
       (*layers_output)(pos, block_name, activations.x.data(), kModelDim);
     }
   }
+
+  // Placeholder for internal test4, do not remove
+
   RMSNormInplace(weights.final_norm_scale.data(), activations.x.data(),
                  kModelDim);
   if (layers_output != nullptr) {