diff --git a/src/main/java/huzpsb/ll4j/model/Model.java b/src/main/java/huzpsb/ll4j/model/Model.java
index 84d21d2..18dc164 100644
--- a/src/main/java/huzpsb/ll4j/model/Model.java
+++ b/src/main/java/huzpsb/ll4j/model/Model.java
@@ -155,6 +155,20 @@ public List<DataEntry> testAndGetWA(DataSet dataSet) {
         return Map.of(((JudgeLayer) layers[layers.length - 1]).result, layers[layers.length - 1].input[((JudgeLayer) layers[layers.length - 1]).result]);
     }
 
+    public double @NotNull [] predictAllResult(double[] input) {
+        AbstractLayer[] layers = this.layers;
+        if (!(layers[layers.length - 1] instanceof JudgeLayer)) {
+            throw new RuntimeException("Last layer is not output layer");
+        }
+        for (AbstractLayer layer : layers) {
+            layer.training = false;
+            layer.input = input;
+            layer.forward();
+            input = layer.output;
+        }
+        return layers[layers.length - 1].input;
+    }
+
     public int predict(DataEntry de) {
         double[] input = de.values;
         for (AbstractLayer layer : layers) {
diff --git a/src/main/kotlin/ltd/guimc/lgzbot/listener/message/MessageFilter.kt b/src/main/kotlin/ltd/guimc/lgzbot/listener/message/MessageFilter.kt
index afe5a22..a0cea1a 100644
--- a/src/main/kotlin/ltd/guimc/lgzbot/listener/message/MessageFilter.kt
+++ b/src/main/kotlin/ltd/guimc/lgzbot/listener/message/MessageFilter.kt
@@ -135,19 +135,16 @@ object MessageFilter {
 
             if (!muted && textMessage.length >= stringLength) {
                 if (LL4JUtils.predict(textMessage)) {
-
                     if (RegexUtils.matchRegexPinyin(adPinyinRegex, textMessage)) {
                         e.group.mute(e.sender, "非法发言内容 (模型预测, 强检查证实)")
                         riskList.add(e.sender)
                         setVl(e.sender.id, 99.0)
                         muted = true
-                    } else {
-                        val botOwner = e.bot.getFriend(Config.BotOwner)
-                        requireNotNull(botOwner)
-                        // botOwner.sendMessage("发现一条模型认为违规的消息, 但正则匹配失败, 请检查.")
-                        // val outputMessage = ForwardMessageBuilder(e.group)
-                        // outputMessage.add(e)
-                        // botOwner.sendMessage(outputMessage.build())
+                    } else if (LL4JUtils.predictMap(textMessage)[1]!! >= 0.9) {
+                        e.sender.mute(120, "非法发言内容 (模型预测认为可能性较高)")
+                        riskList.add(e.sender)
+                        setVl(e.sender.id, 99.0)
+                        muted = true
                     }
                 }
             }