diff --git a/m2cgen/assemblers/boosting.py b/m2cgen/assemblers/boosting.py
index 72eabbca..2265335a 100644
--- a/m2cgen/assemblers/boosting.py
+++ b/m2cgen/assemblers/boosting.py
@@ -2,7 +2,7 @@
 import numpy as np
 
 from m2cgen import ast
-from m2cgen.assemblers import utils
+from m2cgen.assemblers import fallback_expressions, utils
 from m2cgen.assemblers.base import ModelAssembler
 from m2cgen.assemblers.linear import _linear_to_ast
 
@@ -62,7 +62,7 @@ def _assemble_multi_class_output(self, estimator_params):
             for i, e in enumerate(splits)
         ]
 
-        proba_exprs = utils.softmax_exprs(exprs)
+        proba_exprs = fallback_expressions.softmax(exprs)
         return ast.VectorVal(proba_exprs)
 
     def _assemble_bin_class_output(self, estimator_params):
@@ -76,7 +76,7 @@ def _assemble_bin_class_output(self, estimator_params):
         expr = self._assemble_single_output(
             estimator_params, base_score=base_score)
 
-        proba_expr = utils.sigmoid_expr(expr, to_reuse=True)
+        proba_expr = fallback_expressions.sigmoid(expr, to_reuse=True)
 
         return ast.VectorVal([
             ast.BinNumExpr(ast.NumVal(1), proba_expr, ast.BinNumOpType.SUB),
diff --git a/m2cgen/assemblers/fallback_expressions.py b/m2cgen/assemblers/fallback_expressions.py
new file mode 100644
index 00000000..8b8d1b25
--- /dev/null
+++ b/m2cgen/assemblers/fallback_expressions.py
@@ -0,0 +1,50 @@
+"""This module provides an implementation for a variety of functions
+expressed in library's AST.
+
+These AST-based implementations are used as fallbacks in case
+when the target language lacks native support for respective functions
+provided in this module.
+"""
+from m2cgen import ast
+from m2cgen.assemblers import utils
+
+
+def tanh(expr):
+    expr = ast.IdExpr(expr, to_reuse=True)
+    tanh_expr = utils.sub(
+        ast.NumVal(1.0),
+        utils.div(
+            ast.NumVal(2.0),
+            utils.add(
+                ast.ExpExpr(
+                    utils.mul(
+                        ast.NumVal(2.0),
+                        expr)),
+                ast.NumVal(1.0))))
+    return ast.IfExpr(
+        utils.gt(expr, ast.NumVal(44.0)),  # exp(2*x) <= 2^127
+        ast.NumVal(1.0),
+        ast.IfExpr(
+            utils.lt(expr, ast.NumVal(-44.0)),
+            ast.NumVal(-1.0),
+            tanh_expr))
+
+
+def sigmoid(expr, to_reuse=False):
+    neg_expr = ast.BinNumExpr(ast.NumVal(0), expr, ast.BinNumOpType.SUB)
+    exp_expr = ast.ExpExpr(neg_expr)
+    return ast.BinNumExpr(
+        ast.NumVal(1),
+        ast.BinNumExpr(ast.NumVal(1), exp_expr, ast.BinNumOpType.ADD),
+        ast.BinNumOpType.DIV,
+        to_reuse=to_reuse)
+
+
+def softmax(exprs):
+    exp_exprs = [ast.ExpExpr(e, to_reuse=True) for e in exprs]
+    exp_sum_expr = utils.apply_op_to_expressions(
+        ast.BinNumOpType.ADD, *exp_exprs, to_reuse=True)
+    return [
+        ast.BinNumExpr(e, exp_sum_expr, ast.BinNumOpType.DIV)
+        for e in exp_exprs
+    ]
diff --git a/m2cgen/assemblers/utils.py b/m2cgen/assemblers/utils.py
index 0be3b348..ebbbdaf2 100644
--- a/m2cgen/assemblers/utils.py
+++ b/m2cgen/assemblers/utils.py
@@ -19,10 +19,18 @@ def sub(l, r, to_reuse=False):
     return ast.BinNumExpr(l, r, ast.BinNumOpType.SUB, to_reuse=to_reuse)
 
 
+def lt(l, r):
+    return ast.CompExpr(l, r, ast.CompOpType.LT)
+
+
 def lte(l, r):
     return ast.CompExpr(l, r, ast.CompOpType.LTE)
 
 
+def gt(l, r):
+    return ast.CompExpr(l, r, ast.CompOpType.GT)
+
+
 def eq(l, r):
     return ast.CompExpr(l, r, ast.CompOpType.EQ)
 
@@ -79,23 +87,3 @@ def to_2d_array(var):
     else:
         x, y = 1, np.size(var)
     return np.reshape(np.asarray(var), (x, y))
-
-
-def sigmoid_expr(expr, to_reuse=False):
-    neg_expr = ast.BinNumExpr(ast.NumVal(0), expr, ast.BinNumOpType.SUB)
-    exp_expr = ast.ExpExpr(neg_expr)
-    return ast.BinNumExpr(
-        ast.NumVal(1),
-        ast.BinNumExpr(ast.NumVal(1), exp_expr, ast.BinNumOpType.ADD),
-        ast.BinNumOpType.DIV,
-        to_reuse=to_reuse)
-
-
-def softmax_exprs(exprs):
-    exp_exprs = [ast.ExpExpr(e, to_reuse=True) for e in exprs]
-    exp_sum_expr = apply_op_to_expressions(ast.BinNumOpType.ADD, *exp_exprs,
-                                           to_reuse=True)
-    return [
-        ast.BinNumExpr(e, exp_sum_expr, ast.BinNumOpType.DIV)
-        for e in exp_exprs
-    ]
diff --git a/m2cgen/interpreters/dart/tanh.dart b/m2cgen/interpreters/dart/tanh.dart
index f6955237..36d2ae0b 100644
--- a/m2cgen/interpreters/dart/tanh.dart
+++ b/m2cgen/interpreters/dart/tanh.dart
@@ -1,7 +1,28 @@
 double tanh(double x) {
-    if (x > 22.0)
+    // Implementation is taken from
+    // https://github.com/golang/go/blob/master/src/math/tanh.go
+    double z;
+    z = x.abs();
+    if (z > 44.0148459655565271479942397125) {
+        if (x < 0) {
+            return -1.0;
+        }
         return 1.0;
-    if (x < -22.0)
-        return -1.0;
-    return ((exp(2*x) - 1)/(exp(2*x) + 1));
+    }
+    if (z >= 0.625) {
+        z = 1 - 2 / (exp(2 * z) + 1);
+        if (x < 0) {
+            z = -z;
+        }
+        return z;
+    }
+    if (x == 0) {
+        return 0.0;
+    }
+    double s;
+    s = x * x;
+    z = x + x * s
+        * ((-0.964399179425052238628 * s + -99.2877231001918586564) * s + -1614.68768441708447952)
+        / (((s + 112.811678491632931402) * s + 2235.48839060100448583) * s + 4844.06305325125486048);
+    return z;
 }
diff --git a/m2cgen/interpreters/interpreter.py b/m2cgen/interpreters/interpreter.py
index 36f06d6d..4c75ef19 100644
--- a/m2cgen/interpreters/interpreter.py
+++ b/m2cgen/interpreters/interpreter.py
@@ -1,4 +1,5 @@
 from m2cgen import ast
+from m2cgen.assemblers import fallback_expressions
 from m2cgen.interpreters.utils import CachedResult, _get_handler_name
 
 
@@ -120,28 +121,33 @@ def interpret_vector_val(self, expr, **kwargs):
         return self._cg.vector_init(nested)
 
     def interpret_exp_expr(self, expr, **kwargs):
-        assert self.exponent_function_name, "Exponent function is not provided"
+        if self.exponent_function_name is NotImplemented:
+            raise NotImplementedError("Exponent function is not provided")
         self.with_math_module = True
         nested_result = self._do_interpret(expr.expr, **kwargs)
         return self._cg.function_invocation(
             self.exponent_function_name, nested_result)
 
     def interpret_sqrt_expr(self, expr, **kwargs):
-        assert self.sqrt_function_name, "Sqrt function is not provided"
+        if self.sqrt_function_name is NotImplemented:
+            raise NotImplementedError("Sqrt function is not provided")
         self.with_math_module = True
         nested_result = self._do_interpret(expr.expr, **kwargs)
         return self._cg.function_invocation(
             self.sqrt_function_name, nested_result)
 
     def interpret_tanh_expr(self, expr, **kwargs):
-        assert self.tanh_function_name, "Tanh function is not provided"
         self.with_math_module = True
+        if self.tanh_function_name is NotImplemented:
+            return self._do_interpret(
+                fallback_expressions.tanh(expr.expr), **kwargs)
         nested_result = self._do_interpret(expr.expr, **kwargs)
         return self._cg.function_invocation(
             self.tanh_function_name, nested_result)
 
     def interpret_pow_expr(self, expr, **kwargs):
-        assert self.power_function_name, "Power function is not provided"
+        if self.power_function_name is NotImplemented:
+            raise NotImplementedError("Power function is not provided")
         self.with_math_module = True
         base_result = self._do_interpret(expr.base_expr, **kwargs)
         exp_result = self._do_interpret(expr.exp_expr, **kwargs)
diff --git a/m2cgen/interpreters/visual_basic/tanh.bas b/m2cgen/interpreters/visual_basic/tanh.bas
index f9bdbab1..5cd2450d 100644
--- a/m2cgen/interpreters/visual_basic/tanh.bas
+++ b/m2cgen/interpreters/visual_basic/tanh.bas
@@ -1,11 +1,32 @@
 Function Tanh(ByVal number As Double) As Double
-    If number > 44.0 Then  ' exp(2*x) <= 2^127
+    ' Implementation is taken from
+    ' https://github.com/golang/go/blob/master/src/math/tanh.go
+    Dim z As Double
+    z = Math.Abs(number)
+    If z > 44.0148459655565271479942397125 Then
+        If number < 0 Then
+            Tanh = -1.0
+            Exit Function
+        End If
         Tanh = 1.0
         Exit Function
     End If
-    If number < -44.0 Then
-        Tanh = -1.0
+    If z >= 0.625 Then
+        z = 1 - 2 / (Math.Exp(2 * z) + 1)
+        If number < 0 Then
+            z = -z
+        End If
+        Tanh = z
         Exit Function
     End If
-    Tanh = (Math.Exp(2 * number) - 1) / (Math.Exp(2 * number) + 1)
+    If number = 0 Then
+        Tanh = 0.0
+        Exit Function
+    End If
+    Dim s As Double
+    s = number * number
+    z = number + number * s _
+        * ((-0.964399179425052238628 * s + -99.2877231001918586564) * s + -1614.68768441708447952) _
+        / (((s + 112.811678491632931402) * s + 2235.48839060100448583) * s + 4844.06305325125486048)
+    Tanh = z
 End Function
diff --git a/tests/test_fallback_expressions.py b/tests/test_fallback_expressions.py
new file mode 100644
index 00000000..312f19c6
--- /dev/null
+++ b/tests/test_fallback_expressions.py
@@ -0,0 +1,27 @@
+from m2cgen import ast
+from m2cgen.interpreters import PythonInterpreter
+
+from tests.utils import assert_code_equal
+
+
+def test_tanh_fallback_expr():
+    expr = ast.TanhExpr(ast.NumVal(2.0))
+
+    interpreter = PythonInterpreter()
+    interpreter.tanh_function_name = NotImplemented
+
+    expected_code = """
+import math
+def score(input):
+    var1 = 2.0
+    if (var1) > (44.0):
+        var0 = 1.0
+    else:
+        if (var1) < (-44.0):
+            var0 = -1.0
+        else:
+            var0 = (1.0) - ((2.0) / ((math.exp((2.0) * (var1))) + (1.0)))
+    return var0
+"""
+
+    assert_code_equal(interpreter.interpret(expr), expected_code)