diff --git a/constantine/arithmetic/finite_fields.nim b/constantine/arithmetic/finite_fields.nim
index a396a38..d85ce4d 100644
--- a/constantine/arithmetic/finite_fields.nim
+++ b/constantine/arithmetic/finite_fields.nim
@@ -94,10 +94,10 @@ func cswap*(a, b: var FF, ctl: CTBool) {.meter.} =
 #
 # Note: the library currently implements generic routine for odd field modulus.
 #       Routines for special field modulus form:
-#       - Mersenne Prime (2^k - 1),
+#       - Mersenne Prime (2ᵏ - 1),
 #       - Generalized Mersenne Prime (NIST Prime P256: 2^256 - 2^224 + 2^192 + 2^96 - 1)
 #       - Pseudo-Mersenne Prime (2^m - k for example Curve25519: 2^255 - 19)
-#       - Golden Primes (φ^2 - φ - 1 with φ = 2^k for example Ed448-Goldilocks: 2^448 - 2^224 - 1)
+#       - Golden Primes (φ^2 - φ - 1 with φ = 2ᵏ for example Ed448-Goldilocks: 2^448 - 2^224 - 1)
 #       exist and can be implemented with compile-time specialization.
 
 # Note: for `+=`, double, sum
diff --git a/constantine/arithmetic/limbs_montgomery.nim b/constantine/arithmetic/limbs_montgomery.nim
index d88a4d5..de77075 100644
--- a/constantine/arithmetic/limbs_montgomery.nim
+++ b/constantine/arithmetic/limbs_montgomery.nim
@@ -471,8 +471,8 @@ func montyResidue*(r: var Limbs, a, M, r2modM: Limbs,
 # - Apache Milagro Crypto has an alternative implementation
 #   that is more straightforward however:
 #   - the exponent hamming weight is used as loop bounds
-#   - the base^k is stored at each index of a temp table of size k
-#   - the base^k to use is indexed by the hamming weight
+#   - the baseᵏ is stored at each index of a temp table of size k
+#   - the baseᵏ to use is indexed by the hamming weight
 #     of the exponent, leaking this to cache attacks
 #   - in contrast BearSSL touches the whole table to
 #     hide the actual selection
@@ -500,7 +500,7 @@ func montyPowPrologue(
   result = scratchspace.len.getWindowLen()
   # Precompute window content, special case for window = 1
   # (i.e scratchspace has only space for 2 temporaries)
-  # The content scratchspace[2+k] is set at a^k
+  # The content scratchspace[2+k] is set at aᵏ
   # with scratchspace[0] untouched
   if result == 1:
     scratchspace[1] = a
@@ -583,7 +583,7 @@ func montyPow*(
   ## - ``one`` is 1 (mod M) in montgomery representation
   ## - ``m0ninv`` is the montgomery magic constant "-1/M[0] mod 2^WordBitWidth"
   ## - ``scratchspace`` with k the window bitsize of size up to 5
-  ##   This is a buffer that can hold between 2^k + 1 big-ints
+  ##   This is a buffer that can hold between 2ᵏ + 1 big-ints
   ##   A window of of 1-bit (no window optimization) requires only 2 big-ints
   ##
   ## Note that the best window size require benchmarking and is a tradeoff between
diff --git a/constantine/config/curves_parser_field.nim b/constantine/config/curves_parser_field.nim
index 118aa30..70b0c5f 100644
--- a/constantine/config/curves_parser_field.nim
+++ b/constantine/config/curves_parser_field.nim
@@ -59,7 +59,7 @@ type
     ## denoted `E(𝔽p)` in Short Weierstrass form
     ##   y² = x³ + Ax + B
     ##
-    ## If E(𝔽p^k), the elliptic curve defined over the extension field
+    ## If E(𝔽pᵏ), the elliptic curve defined over the extension field
     ## of degree k, the embedding degree, admits an isomorphism
     ## to a curve E'(Fp^(k/d)), we call E' a twisted curve.
     ##
@@ -67,7 +67,7 @@ type
     ##   y² = x³ + Ax/µ² + B/µ³ for a D-Twist (Divisor)
     ## or
     ##   y² = x³ + µ²Ax + µ³B for a M-Twist (Multiplicand)
-    ## with the polynomial x^k - µ being irreducible.
+    ## with the polynomial xᵏ - µ being irreducible.
     ##
     ## i.e. if d == 2, E' is a quadratic twist and µ is a quadratic non-residue
     ## if d == 4, E' is a quartic twist
diff --git a/constantine/config/precompute.nim b/constantine/config/precompute.nim
index c77c763..d69fea5 100644
--- a/constantine/config/precompute.nim
+++ b/constantine/config/precompute.nim
@@ -276,7 +276,7 @@ func invModBitwidth[T: SomeUnsignedInt](a: T): T =
   # For a and m to be coprimes, a must be odd.
   #
   # We have the following relation
-  # ax ≡ 1 (mod 2^k) <=> ax(2 - ax) ≡ 1 (mod 2^(2k))
+  # ax ≡ 1 (mod 2ᵏ) <=> ax(2 - ax) ≡ 1 (mod 2^(2k))
   # which grows in O(log(log(a)))
   checkOdd(a)
 
diff --git a/constantine/isogeny/frobenius.nim b/constantine/isogeny/frobenius.nim
index e7f32a7..5b5cb7e 100644
--- a/constantine/isogeny/frobenius.nim
+++ b/constantine/isogeny/frobenius.nim
@@ -37,13 +37,13 @@ import
 # or          √-2 or √-5
 
 func frobenius_map*(r: var Fp, a: Fp, k: static int = 1) {.inline.} =
-  ## Computes a^(p^k)
+  ## Computes a^(pᵏ)
   ## The p-power frobenius automorphism on 𝔽p
   ## This is identity per Fermat's little theorem
   r = a
 
 func frobenius_map*(r: var Fp2, a: Fp2, k: static int = 1) {.inline.} =
-  ## Computes a^(p^k)
+  ## Computes a^(pᵏ)
   ## The p-power frobenius automorphism on 𝔽p2
   when (k and 1) == 1:
     r.conj(a)
@@ -54,14 +54,14 @@ func frobenius_map*(r: var Fp2, a: Fp2, k: static int = 1) {.inline.} =
 # -----------------------------------------------------------------
 
 func frobenius_map*[C](r: var Fp4[C], a: Fp4[C], k: static int = 1) {.inline.} =
-  ## Computes a^(p^k)
+  ## Computes a^(pᵏ)
   ## The p-power frobenius automorphism on 𝔽p4
   r.c0.frobenius_map(a.c0, k)
   r.c1.frobenius_map(a.c1, k)
   r.c1.mulCheckSparse frobMapConst(C, 3, k)
 
 func frobenius_map*[C](r: var Fp6[C], a: Fp6[C], k: static int = 1) {.inline.} =
-  ## Computes a^(p^k)
+  ## Computes a^(pᵏ)
   ## The p-power frobenius automorphism on 𝔽p6
   r.c0.frobenius_map(a.c0, k)
   r.c1.frobenius_map(a.c1, k)
@@ -77,7 +77,7 @@ func frobenius_map*[C](r: var Fp6[C], a: Fp6[C], k: static int = 1) {.inline.} =
     {.error: "Not Implemented".}
 
 func frobenius_map*[C](r: var Fp12[C], a: Fp12[C], k: static int = 1) {.inline.} =
-  ## Computes a^(p^k)
+  ## Computes a^(pᵏ)
   ## The p-power frobenius automorphism on 𝔽p12
   static: doAssert r.c0 is Fp4
   staticFor i, 0, r.coords.len:
diff --git a/constantine/pairing/lines_common.nim b/constantine/pairing/lines_common.nim
index dce8a4b..48c2722 100644
--- a/constantine/pairing/lines_common.nim
+++ b/constantine/pairing/lines_common.nim
@@ -19,7 +19,7 @@ import
 
 type
   Line*[F] = object
-    ## Packed line representation over a E'(Fp^k/d)
+    ## Packed line representation over a E'(Fpᵏ/d)
     ## with k the embedding degree and d the twist degree
     ## i.e. for a curve with embedding degree 12 and sextic twist
     ## F is Fp2
diff --git a/constantine/tower_field_extensions/README.md b/constantine/tower_field_extensions/README.md
index 7b5927b..b480872 100644
--- a/constantine/tower_field_extensions/README.md
+++ b/constantine/tower_field_extensions/README.md
@@ -24,7 +24,7 @@ From Ben Edgington, https://hackmd.io/@benjaminion/bls12-381
 >
 > Applying our rule, by substituting $x^2 = -1$, gives us the final result $(a, b) \times (c, d) =$$ac + (ad+bc)x + bdx^2 =$$(ac-bd) + (ad+bc)x =$$(ac-bd, ad+bc)$. This might look a little familiar from complex arithmetic: $(a+ib) \times (c+id) =$$(ac-bd) + (ad+bc)i$. This is not a coincidence! The complex numbers are a quadratic extension of the real numbers.
 >
-> Complex numbers can't be extended any further because there are [no irreducible polynomials over the complex numbers](https://en.wikipedia.org/wiki/Fundamental_theorem_of_algebra). But for finite fields, if we can find an irreducible $k$-degree polynomial in our field $F_q$, and we often can, then we are able to extend the field to $F_{q^k}$, and represent the elements of the extended field as degree $k-1$ polynomials, $a_0 + a_1x +$$...$$+ a_{k-1}x^{k-1}$. We can represent this compactly as $(a_0,...,a_{k-1})$, as long as we remember that there may be some very funky arithmetic going on.
+> Complex numbers can't be extended any further because there are [no irreducible polynomials over the complex numbers](https://en.wikipedia.org/wiki/Fundamental_theorem_of_algebra). But for finite fields, if we can find an irreducible $k$-degree polynomial in our field $F_q$, and we often can, then we are able to extend the field to $F_{qᵏ}$, and represent the elements of the extended field as degree $k-1$ polynomials, $a_0 + a_1x +$$...$$+ a_{k-1}x^{k-1}$. We can represent this compactly as $(a_0,...,a_{k-1})$, as long as we remember that there may be some very funky arithmetic going on.
 >
 > Also worth noting is that modular reductions like this (our reduction rule) can be chosen so that they play nicely with the twisting operation.
 >
diff --git a/docs/optimizations.md b/docs/optimizations.md
index e139436..3db6c2f 100644
--- a/docs/optimizations.md
+++ b/docs/optimizations.md
@@ -38,10 +38,10 @@ The optimizations can be of algebraic, algorithmic or "implementation details" n
   - [x] Montgomery Representation
   - [ ] Barret Reduction
   - [ ] Unsaturated Representation
-    - [ ] Mersenne Prime (2^k - 1),
+    - [ ] Mersenne Prime (2ᵏ - 1),
     - [ ] Generalized Mersenne Prime (NIST Prime P256: 2^256 - 2^224 + 2^192 + 2^96 - 1)
     - [ ] Pseudo-Mersenne Prime (2^m - k for example Curve25519: 2^255 - 19)
-    - [ ] Golden Primes (φ^2 - φ - 1 with φ = 2^k for example Ed448-Goldilocks: 2^448 - 2^224 - 1)
+    - [ ] Golden Primes (φ^2 - φ - 1 with φ = 2ᵏ for example Ed448-Goldilocks: 2^448 - 2^224 - 1)
     - [ ] any prime modulus (lazy carry)
 
 - Montgomery Reduction
diff --git a/sage/derive_pairing.sage b/sage/derive_pairing.sage
index 7df2ad1..e4ddf9b 100644
--- a/sage/derive_pairing.sage
+++ b/sage/derive_pairing.sage
@@ -198,7 +198,7 @@ def genFinalExp(curve_name, curve_config):
     scale = 3*(u^3-u^2+1)
     scaleDesc = ' * 3*(u^3-u^2+1)'
 
-  fexp = (p^k - 1)//r
+  fexp = (pᵏ - 1)//r
   fexp *= scale
 
   buf = f'const {curve_name}_pairing_finalexponent* = block:\n'
diff --git a/tests/t_fp12_frobenius.nim b/tests/t_fp12_frobenius.nim
index b0d6126..b9407e0 100644
--- a/tests/t_fp12_frobenius.nim
+++ b/tests/t_fp12_frobenius.nim
@@ -25,5 +25,5 @@ runFrobeniusTowerTests(
   Iters = 8,
   TestCurves = TestCurves,
   moduleName = "test_fp12_frobenius",
-  testSuiteDesc = "𝔽p12 Frobenius map: Frobenius(a, k) = a^(p^k) (mod p^12)"
+  testSuiteDesc = "𝔽p12 Frobenius map: Frobenius(a, k) = a^(pᵏ) (mod p^12)"
 )
diff --git a/tests/t_fp2_frobenius.nim b/tests/t_fp2_frobenius.nim
index cc37f49..5d16237 100644
--- a/tests/t_fp2_frobenius.nim
+++ b/tests/t_fp2_frobenius.nim
@@ -26,5 +26,5 @@ runFrobeniusTowerTests(
   Iters = 8,
   TestCurves = TestCurves,
   moduleName = "test_fp2_frobenius",
-  testSuiteDesc = "𝔽p2 Frobenius map: Frobenius(a, k) = a^(p^k) (mod p²)"
+  testSuiteDesc = "𝔽p2 Frobenius map: Frobenius(a, k) = a^(pᵏ) (mod p²)"
 )
diff --git a/tests/t_fp4_frobenius.nim b/tests/t_fp4_frobenius.nim
index fd505ab..812ce92 100644
--- a/tests/t_fp4_frobenius.nim
+++ b/tests/t_fp4_frobenius.nim
@@ -25,5 +25,5 @@ runFrobeniusTowerTests(
   Iters = 8,
   TestCurves = TestCurves,
   moduleName = "test_fp4_frobenius",
-  testSuiteDesc = "𝔽p4 Frobenius map: Frobenius(a, k) = a^(p^k) (mod p⁴)"
+  testSuiteDesc = "𝔽p4 Frobenius map: Frobenius(a, k) = a^(pᵏ) (mod p⁴)"
 )
diff --git a/tests/t_fp6_frobenius.nim b/tests/t_fp6_frobenius.nim
index cc9701f..a7c5423 100644
--- a/tests/t_fp6_frobenius.nim
+++ b/tests/t_fp6_frobenius.nim
@@ -26,5 +26,5 @@ runFrobeniusTowerTests(
   Iters = 8,
   TestCurves = TestCurves,
   moduleName = "test_fp6_frobenius",
-  testSuiteDesc = "𝔽p6 Frobenius map: Frobenius(a, k) = a^(p^k) (mod p⁶)"
+  testSuiteDesc = "𝔽p6 Frobenius map: Frobenius(a, k) = a^(pᵏ) (mod p⁶)"
 )