diff --git a/be/src/exec/pipeline/scan/scan_operator.cpp b/be/src/exec/pipeline/scan/scan_operator.cpp
index 312e7f09ef963..51a6dc77b0398 100644
--- a/be/src/exec/pipeline/scan/scan_operator.cpp
+++ b/be/src/exec/pipeline/scan/scan_operator.cpp
@@ -233,7 +233,9 @@ StatusOr<ChunkPtr> ScanOperator::pull_chunk(RuntimeState* state) {
         begin_pull_chunk(res);
         // for query cache mechanism, we should emit EOS chunk when we receive the last chunk.
         auto [tablet_id, is_eos] = _should_emit_eos(res);
-        eval_runtime_bloom_filters(res.get());
+        if (!_scan_node->support_push_down_runtime_filter_to_reader()) {
+            eval_runtime_bloom_filters(res.get());
+        }
         res->owner_info().set_owner_id(tablet_id, is_eos);
     }
 
diff --git a/be/src/exec/pipeline/scan/scan_operator.h b/be/src/exec/pipeline/scan/scan_operator.h
index 24405ced9ec15..4c6e833a34f50 100644
--- a/be/src/exec/pipeline/scan/scan_operator.h
+++ b/be/src/exec/pipeline/scan/scan_operator.h
@@ -14,6 +14,7 @@
 
 #pragma once
 
+#include "exec/scan_node.h"
 #include "exec/pipeline/source_operator.h"
 #include "exec/query_cache/cache_operator.h"
 #include "exec/query_cache/lane_arbiter.h"
@@ -90,7 +91,7 @@ class ScanOperator : public SourceOperator {
 
     virtual int64_t get_scan_table_id() const { return -1; }
 
-    virtual bool support_push_down_runtime_filter_to_reader() const { return _scan_node->is_push_down_runtime_filter_to_reader(); }
+    virtual bool support_push_down_runtime_filter_to_reader() const { return _scan_node->support_push_down_runtime_filter_to_reader(); }
 
 protected:
     static constexpr size_t kIOTaskBatchSize = 64;
diff --git a/be/src/formats/parquet/group_reader.cpp b/be/src/formats/parquet/group_reader.cpp
index ca7cded573f11..0cd392970c4e4 100644
--- a/be/src/formats/parquet/group_reader.cpp
+++ b/be/src/formats/parquet/group_reader.cpp
@@ -117,7 +117,7 @@ Status GroupReader::get_next(ChunkPtr* chunk, size_t* row_count) {
         has_filter = true;
     }
 
-    if (SIMD::contain_nonzero(chunk_filter) && _param.runtime_filter_collector) {
+    if (chunk_size > 0 && _param.runtime_filter_collector) {
         // bloom filter
         for (auto &it: _runtime_filter_by_slot) {
             const JoinRuntimeFilter *filter = it.second;
@@ -267,7 +267,7 @@ void GroupReader::_process_columns_and_conjunct_ctxs() {
             // cause of runtime filter min/max have been added to conjunct_ctxs_by_slot
             // so just check runtime filter exist and add to dict_filter_ctx
             if (_runtime_filter_by_slot.find(slot_id) != _runtime_filter_by_slot.end()) {
-                _dict_filter_ctx.use_as_dict_filter_column(read_col_idx, slot_id, _runtime_filter_by_slot.at(slot_id));
+                _dict_filter_ctx.use_as_dict_filter_column(slot_id, _runtime_filter_by_slot.at(slot_id));
             }
         } else {
             bool has_conjunct = false;
@@ -618,9 +618,9 @@ Status GroupReader::DictFilterContext::rewrite_conjunct_ctxs_to_predicates(
             ctx.use_merged_selection = false;
             ctx.merged_selection.assign(dict_value_column->size(), 1);
             join_filter->compute_hash({dict_value_column.get()}, &ctx);
-            filter->evaluate(dict_value_column.get(), &ctx);
+            join_filter->evaluate(dict_value_column.get(), &ctx);
             bool all_zero = false;
-            ColumnHelper::merge_two_filters(filter, ctx.merged_selection.data(), &all_zero);
+            ColumnHelper::merge_two_filters(&filter, ctx.merged_selection.data(), &all_zero);
             if (all_zero) {
                 *is_group_filtered = true;
                 return Status::OK();
diff --git a/be/src/formats/parquet/group_reader.h b/be/src/formats/parquet/group_reader.h
index 2eb5e0673377f..63201e0f26add 100644
--- a/be/src/formats/parquet/group_reader.h
+++ b/be/src/formats/parquet/group_reader.h
@@ -165,7 +165,7 @@ class GroupReader {
     // conjunct ctxs that eval after chunk is dict decoded
     std::vector<ExprContext*> _left_conjunct_ctxs;
 
-    std::unordered_map<SlotId, JoinRuntimeFilter*> _runtime_filter_by_slot;
+    std::unordered_map<SlotId, const JoinRuntimeFilter*> _runtime_filter_by_slot;
 
     // active columns that hold read_col index
     std::vector<int> _active_column_indices;