diff --git a/egs/librispeech/ASR/pruned_transducer_stateless7/conformer.py b/egs/librispeech/ASR/pruned_transducer_stateless7/conformer.py
index a0cc8fc1f..944e010e8 100644
--- a/egs/librispeech/ASR/pruned_transducer_stateless7/conformer.py
+++ b/egs/librispeech/ASR/pruned_transducer_stateless7/conformer.py
@@ -297,9 +297,10 @@ class ConformerEncoderLayer(nn.Module):
             min_positive=0.45, max_positive=0.55,
             max_abs=6.0,
         )
-        self.max_eig = MaxEig(
-            d_model, channel_dim=-1,
-        )
+        self.whiten = Whiten(num_groups=1,
+                             whitening_limit=5.0,
+                             prob=(0.025, 0.25),
+                             grad_scale=0.01)
 
 
     def forward(
@@ -352,7 +353,7 @@ class ConformerEncoderLayer(nn.Module):
 
         src = src + self.feed_forward3(src)
 
-        src = self.norm_final(self.max_eig(self.balancer(src)))
+        src = self.norm_final(self.balancer(src))
 
         delta = src - src_orig
         bypass_scale = self.bypass_scale
@@ -360,7 +361,7 @@ class ConformerEncoderLayer(nn.Module):
             bypass_scale = bypass_scale.clamp(min=0.1, max=1.0)
         src = src_orig + delta * self.bypass_scale
 
-        return src
+        return self.whiten(src)
 
 
 class ConformerEncoder(nn.Module):