diff --git a/predtuner/approxapp.py b/predtuner/approxapp.py
index 61d203b41ceaf9e39200e7112bc16851cb758fdc..5cd82a19e770f6c34208fac1bda6d738c5f6ed4c 100644
--- a/predtuner/approxapp.py
+++ b/predtuner/approxapp.py
@@ -89,6 +89,13 @@ class ApproxApp(abc.ABC):
                 knobs.append(baseline_knob)
         return baseline_knob
 
+    def add_baseline_to_knobs(self, approxes: KnobsT):
+        approxes = approxes.copy()
+        for op_name in self.ops:
+            if op_name not in approxes:
+                approxes[op_name] = self.baseline_knob.name
+        return approxes
+
 
 class BaselineKnob(ApproxKnob):
     def __init__(self, name: str = "__baseline__"):
diff --git a/predtuner/modeledapp.py b/predtuner/modeledapp.py
index 6c5fe7008b41c9ea7859ada00646405512700c9d..06efc8cc69e7898b667f775bb68bb1a5be60a269 100644
--- a/predtuner/modeledapp.py
+++ b/predtuner/modeledapp.py
@@ -144,12 +144,16 @@ class LinearPerfModel(IPerfModel):
     """Weighted linear performance predictor based on cost of each operator."""
 
     def __init__(
-        self, op_costs: Dict[str, float], knob_speedups: Dict[str, float]
+        self,
+        app: ModeledApp,
+        op_costs: Dict[str, float],
+        knob_speedups: Dict[str, float],
     ) -> None:
         import numpy as np
         import pandas as pd
 
         super().__init__()
+        self.app = app
         knob_cost_factor_v = 1 / np.array(list(knob_speedups.values()))
         layer_cost_v = np.array(list(op_costs.values()))
         costs = np.outer(layer_cost_v, knob_cost_factor_v)
@@ -163,6 +167,7 @@ class LinearPerfModel(IPerfModel):
 
     def measure_perf(self, with_approxes: KnobsT) -> float:
         """We implement this using a weighted linear performance model."""
+        with_approxes = self.app.add_baseline_to_knobs(with_approxes)
         return float(
             sum(self.cost_df.loc[layer, knob] for layer, knob in with_approxes.items())
         )
@@ -204,6 +209,7 @@ class QoSModelP1(IQoSModel):
     def measure_qos(self, with_approxes: KnobsT) -> float:
         """Implementation of model."""
         assert self.baseline_tensor is not None
+        with_approxes = self.app.add_baseline_to_knobs(with_approxes)
         delta_tensors = np.array(
             [self.delta_tensors[op][knob] for op, knob in with_approxes.items()]
         )
@@ -274,6 +280,7 @@ class QoSModelP2(IQoSModel):
 
     def measure_qos(self, with_approxes: KnobsT) -> float:
         assert self.baseline_qos is not None and self.qos_df is not None
+        with_approxes = self.app.add_baseline_to_knobs(with_approxes)
         delta_qoses = (
             np.array([self.qos_df.loc[kv] for kv in with_approxes.items()])
             - self.baseline_qos
diff --git a/predtuner/torchapp.py b/predtuner/torchapp.py
index c93bd8adc9e1163fde6603e1734d9764bcbc7964..0e6fc287be85cc0b73766ce91c1034a6e07186d1 100644
--- a/predtuner/torchapp.py
+++ b/predtuner/torchapp.py
@@ -1,6 +1,6 @@
 import abc
 from pathlib import Path
-from typing import Any, Callable, Dict, List, Optional, Set, Tuple, Union
+from typing import Any, Callable, List, Optional, Set, Tuple, Union
 
 import numpy as np
 import torch
@@ -152,7 +152,7 @@ class TorchApp(ModeledApp, abc.ABC):
         p1_storage = self.model_storage / "p1.pkl" if self.model_storage else None
         p2_storage = self.model_storage / "p2.json" if self.model_storage else None
         return [
-            LinearPerfModel(self._op_costs, self._knob_speedups),
+            LinearPerfModel(self, self._op_costs, self._knob_speedups),
             QoSModelP1(
                 self, self._get_raw_output_valset, batched_valset_qos, p1_storage
             ),