diff --git a/common/sampling.cpp b/common/sampling.cpp
index 9c707a5bb9..9813762eca 100644
--- a/common/sampling.cpp
+++ b/common/sampling.cpp
@@ -129,7 +129,7 @@ struct common_sampler {
             const uint32_t sampled_probs_count = llama_get_backend_sampled_probs_count_ith(ctx, idx);
             cur.reserve(sampled_probs_count);
             for (uint32_t i = 0; i < sampled_probs_count; ++i) {
-                cur.emplace_back(llama_token_data{sampled_ids[i], 0.0f, sampled_probs[i]});
+                cur.emplace_back(llama_token_data{sampled_ids[i], sampled_logits[i], sampled_probs[i]});
             }
         } else if (sampled_logits) {
             const uint32_t sampled_logits_count = llama_get_backend_sampled_logits_count_ith(ctx, idx);
diff --git a/src/llama-sampling.cpp b/src/llama-sampling.cpp
index 456e050201..2cffa524cd 100644
--- a/src/llama-sampling.cpp
+++ b/src/llama-sampling.cpp
@@ -461,7 +461,7 @@ llama_token llama_sampler_sample(struct llama_sampler * smpl, struct llama_conte
         const uint32_t sampled_probs_count = llama_get_backend_sampled_probs_count_ith(ctx, idx);
         cur.reserve(sampled_probs_count);
         for (uint32_t i = 0; i < sampled_probs_count; ++i) {
-            cur.emplace_back(llama_token_data{sampled_ids[i], 0.0f, sampled_probs[i]});
+            cur.emplace_back(llama_token_data{sampled_ids[i], sampled_logits[i], sampled_probs[i]});
         }
     } else if (sampled_logits) {
         const uint32_t sampled_logits_count = llama_get_backend_sampled_logits_count_ith(ctx, idx);