A Deeper Look at
Machine Learning-Based Cryptanalysis

Adrien Benamira1, David Gerault1,2, Thomas Peyrin1, and Quan Quan Tan1
1 Nanyang Technological University, Singapore
2 University of Surrey, UK

adrien002@e.ntu.edu.sg, dagerault@gmail.com, thomas.peyrin@ntu.edu.sg,

quanquan001@e.ntu.edu.sg

Keywords: Diﬀerential Cryptanalysis, SPECK, Machine Learning, Deep Neu-
ral Networks, Interpretability

Abstract. At CRYPTO’19, Gohr proposed a new cryptanalysis strat-
egy based on the utilisation of machine learning algorithms. Using deep
neural networks, he managed to build a neural based distinguisher that
surprisingly surpassed state-of-the-art cryptanalysis eﬀorts on one of the
versions of the well studied NSA block cipher SPECK (this distinguisher
could in turn be placed in a larger key recovery attack). While this
work opens new possibilities for machine learning-aided cryptanalysis, it
remains unclear how this distinguisher actually works and what infor-
mation is the machine learning algorithm deducing. The attacker is left
with a black-box that does not tell much about the nature of the possible
weaknesses of the algorithm tested, while hope is thin as interpretability
of deep neural networks is a well-known diﬃcult task.
In this article, we propose a detailed analysis and thorough explanations
of the inherent workings of this new neural distinguisher. First, we stud-
ied the classiﬁed sets and tried to ﬁnd some patterns that could guide
us to better understand Gohr’s results. We show with experiments that
the neural distinguisher generally relies on the diﬀerential distribution
on the ciphertext pairs, but also on the diﬀerential distribution in penul-
timate and antepenultimate rounds. In order to validate our ﬁndings, we
construct a distinguisher for SPECK cipher based on pure cryptanaly-
sis, without using any neural network, that achieves basically the same
accuracy as Gohr’s neural distinguisher and with the same eﬃciency
(therefore improving over previous non-neural based distinguishers).
Moreover, as another approach, we provide a machine learning-based
distinguisher that strips down Gohr’s deep neural network to a bare
minimum. We are able to remain very close to Gohr’s distinguishers’
accuracy using simple standard machine learning tools. In particular, we
show that Gohr’s neural distinguisher is in fact inherently building a
very good approximation of the Diﬀerential Distribution Table (DDT)
of the cipher during the learning phase, and using that information to
directly classify ciphertext pairs. This result allows a full interpretability
of the distinguisher and represents on its own an interesting contribution
towards interpretability of deep neural networks.
Finally, we propose some method to improve over Gohr’s work and possi-
ble new neural distinguishers settings. All our results are conﬁrmed with

experiments we have been conducted on SPECK block cipher (source
code available online).

Introduction

While modern symmetric-key cryptography designs are heavily relying on se-
curity by construction with strong security arguments (resistance against sim-
ple diﬀerential/linear attacks, study of algebraic properties, etc.), cryptanalysis
remains a crucial part of a cipher’s validation process. Only a primitive that
went through active and thorough scrutiny of third-party cryptanalysts should
gain enough trust by the community to be considered as secure. However, there
has been more and more cipher proposals in the past decade (especially with
the recent rise of lightweight cryptography) and cryptanalysis eﬀort could not
really keep up the pace: conducting cryptanalysis remains a very tough and
low-rewarding task.

In order to partially overcome this shortage in cryptanalysts manpower, a
recent trend arose of automating as much as possible the various tasks of an
attacker. Typically, searching for diﬀerential and linear characteristics can now
be modeled as Satisﬁability/Satisﬁability Modulo Theories [17] (SAT/SMT),
Mixed Linear Integer Programming [18] (MILP) or Constraint Programming [25]
(CP) problems, which can in turn simply be handled by an appropriate solver.
The task of the cryptanalyst is therefore reduced to only providing an eﬃcient
modeling of the problem to be studied. Due to the impressive results considering
the simplicity of the process, a lot of advances have been made in the past decade
in this very active research ﬁeld and this even improved the ciphers designs
themselves (how to choose better cryptographic bricks and how to assemble
them has been made much easier thanks to these new automated tools). One
is then naturally tempted to push this idea further by even getting rid of the
modeling part. More generally, can a tool recognize possible weaknesses/patterns
in a cipher by just interacting with it, with as little input as possible from the
cryptanalysts? One does not expect such a tool to replace a cryptanalyst’s job,
but it might come in handy for easily pre-checking a cipher (or reduced versions
of it) for possible weaknesses.

Machine learning and particularly deep learning have recently attracted a
lot of attention, due to impressive advances in important research areas such
as computer vision, speech recognition, etc. Some possible connections between
cryptography and machine learning were already identiﬁed in [21] and we have
seen many applications of machine learning for side-channels analysis [16]. How-
ever, machine learning for black-box cryptanalysis remained mostly unexplored
until Gohr’s article presented at CRYPTO’19 [11].

In his work, Gohr trained a deep neural network on labeled data composed
of ciphertext pairs: half the data coming from ciphering plaintexts pairs with a
ﬁxed input diﬀerence with the cipher studied, half from random values. He then
checks if the trained neural network is able to classify accurately random from
real ciphertext pairs. Quite surprisingly, when applying his framework to the

block cipher SPECK-32/64 (the 32-bit block 64-bit key version of SPECK [2]),
he managed to obtain a good accuracy for a non-negligible number of rounds.
He even managed to mount a key recovery process on top of his neural distin-
guisher, eventually leading to the current best known key recovery attack for
this number of rounds (improving over works on SPECK-32/64 such as [6, 24]).
Even if his distinguisher/key recovery attack had not been improving over the
state-of-the-art, the prospect of a generic tool that could pre-scan for vulnera-
bilities in a cryptographic primitive (while reaching an accuracy close to exiting
cryptanalysis) would have been very attractive anyway.

Yet, Gohr’s paper actually opened many questions. The most important,
listed by the author as an open problem, is the interpretability of the distin-
guisher. An obvious issue with a neural distinguisher is that its black-box nature
is not really telling us much about the actual weakness of the cipher analyzed.
More generally, interpretability for deep neural networks has been known to be
a very complex problem and represents a key challenge for the machine learning
community. At ﬁrst sight, it seems therefore very diﬃcult to make any advances
in this direction.

Another interesting aspect to explore is to try to match Gohr’s neural dis-
tinguisher/key recovery attack with classical cryptanalysis tools. It remains very
surprising that a trained deep neural network can perform better than the
scrutiny of experienced cryptanalysts. As remarked by Gohr, his neural dis-
tinguisher is mostly diﬀerential in nature (on the ciphertext pairs), but some
unknown extra property is exploited. Indeed, as demonstrated by one of his ex-
periments, the neural distinguisher can still distinguish between a real and a
random set that have the exact same diﬀerential distribution on the ciphertext
pairs. Since we know there is some property that researchers have not seen or
exploited, what is it?

Finally, a last natural question is: can we do better? Are there some better

settings that could improve the accuracy of Gohr’s distinguishers?

Our Contributions.
In this article, we analyze the behavior of Gohr’s neural
distinguishers when working on SPECK-32/64 cipher. We ﬁrst study in detail
the classiﬁed sets of real/random ciphertext pairs in order to get some hints on
what criterion the neural network is actually basing its decisions on. Looking for
patterns, we observe that the neural distinguisher is very probably deducing some
diﬀerential conditions not on the ciphertext pairs directly, but on the penultimate
or antepenultimate rounds. We then conduct some experiments to validate our
hypothesis.

In order to further conﬁrm our ﬁndings, we construct for 5, 6 and 7-round
reduced SPECK-32/64 a new distinguisher purely based on cryptanalysis, with-
out any neural network or machine learning algorithm, that matches Gohr’s
neural distinguisher’s accuracy while actually being faster and using the same
amount of precomputation/training data. In short, our distinguisher relies on
selective partial decryption: in order to attack nr rounds, some hypothesis is
made on some bits of the last round subkey and partial decryption is performed,
eventually ﬁltered by a precomputed approximated DDT on nr − 1 rounds.

We then take a diﬀerent approach by tackling the problem not from the crypt-
analysis side, but the machine learning side. More precisely, as a deep learning
model learns high-level features by itself, in order to reach full interpretability
we need to discover what these features are. By analyzing the components of
Gohr’s neural network, we managed to identify a procedure to model these fea-
tures, while retaining almost the same accuracy as Gohr’s neural distinguishers.
Moreover, we also show that our method performs similarly on other primitives
by applying it on the SIMON block cipher. This result is interesting from a cryp-
tography perspective, but also from a machine learning perspective, showing an
example of interpretability by transformation of a deep neural network.

Finally, we explore possible improvements over Gohr’s neural distinguishers.
By using batches of ciphertexts instead of pairs, we are able to signiﬁcantly im-
prove the accuracy of the distinguisher, while maintaining identical experimental
conditions.

Outline.
In Section 2, we introduce notations as well as basic cryptanalysis
and machine learning concepts that will be used in the rest of the paper. In
Section 3, we describe in more detail the various experiments conducted by
Gohr and the corresponding results. We provide in Section 4 an explanation of
his neural distinguishers as well as the description of an actual cryptanalysis-only
distinguisher that matches Gohr’s accuracy. We propose in Section 5 a machine
learning approach to enable interpretability of the neural distinguishers. Finally,
we studied possible improvements in Section 6.

2 Preliminaries

In the rest of this article, ⊕, ∧ and (cid:1) will denote the
Basic notations.
eXclusive-OR operation, the bitwise AND operation and the modular addition3
respectively. A right/left bit rotation will be denoted as ≫ and ≪ respectively,
while a||b will represent the concatenation of two bit strings a and b.

2.1 A Brief Description of SPECK

The lightweight family of ARX block ciphers SPECK was proposed by the US
National Security Agency (NSA) [2] in 2013, targeting mainly good performances
on micro-controllers. Several versions of the cipher have been proposed within
its family, but in this article (and in Gohr’s work [11]) we will focus mainly on
SPECK-32/64, the 32-bit block 64-bit key version of SPECK, which is composed
of 22 rounds (for simplicity, SPECK-32/64 will be referred to as SPECK in the
rest of the article).

The 32-bit internal state is divided into a 16-bit left and a 16-bit right part,
that we will generally denote li and ri at round i respectively, and is initialised
with the plaintext (l0||r0) ← P . The round function of the cipher is then a very
3 The modulo will be stated explicitly if it is not clear from the context

simple Feistel structure combining bitwise XOR operation and 16-bit modular
addition. See Figure 1 where ki represents the 16-bit subkey at round i and
where α = 7, β = 2. The ﬁnal ciphertext C is then obtained as C ← (l22||r22).
The subkeys are generated with a key schedule that is very similar to the round
function (we refer to [2] for a complete description, as we do not make use of the
details of the key schedule in this article).

li+1 = ((li ≫ α) (cid:1) ri) ⊕ ki
ri+1 = (ri ≪ β) ⊕ li+1

≫ α

≪ β

li+1

ri+1

Fig. 1: The SPECK-32/64 round function.

2.2 Diﬀerential Cryptanalysis

Diﬀerential cryptanalysis studies the propagation of a diﬀerence through a ci-
2 and x, x(cid:48) be two diﬀerent inputs for f with a
pher. Let a function f : Fb
diﬀerence ∆x = x⊕x(cid:48). Let y = f (x) and y(cid:48) = f (x(cid:48)) and a diﬀerence ∆y = y⊕y(cid:48).
f−→ ∆y):
Then, we are interested in the transition probability from ∆x to ∆y (∆x

2 → Fb

P(∆x

f−→ ∆y) :=

#{x|f (x) ⊕ f (x ⊕ ∆x) = ∆y}

One classical tool for diﬀerential cryptanalysis is the Diﬀerence Distribution Ta-
ble (DDT), which simply lists the diﬀerential transition probabilities for each
possible input/output diﬀerence pairs (∆x, ∆y). The studied function f is usu-
ally some Sbox, or some small cipher sub-component, as the DDT of an entire
64-bit or 128-bit cipher would obviously be too large to store.

Since SPECK is internally composed of a left and right part, for a ciphertext
C we will denote by Cl and Cr its 16-bit left and right parts respectively. Then,
for two ciphertexts C and C(cid:48), we will denote ∆L the XOR diﬀerence Cl ⊕ C(cid:48)
between the left parts of the two ciphertexts (respectively ∆R = Cr ⊕ C(cid:48)
l
r for the
right parts). Moreover, for a round i, we will denote by Vi the diﬀerence between
the two parts of the internal state Vi = li ⊕ ri.

2.3 Deep Neural Networks

Deep Neural Networks (DNN) are a family of non-linear machine learning clas-
siﬁers that have gained popularity since their success in addressing a variety of
data-driven tasks, such as computer vision, speech recognition, etc.

n(cid:88)

The main problem tackled by DNN is, given a dataset D = {(x0, y0)...(xn, yn)},
with xi ∈ X being samples and yi ∈ [0, . . . , l] being labels, to ﬁnd the optimal
parameters θ∗ for the DN Nθ model, with the parameters θ such that:

θ∗ = argmin

L(yi, DN N θ(xi))

(1)
with L being the loss function. As there is no literal expression of θ∗, the ap-
proximate solution will depend on the chosen optimization algorithm such as the
stochastic gradient descent. Moreover, hyper-parameters of the problem (param-
eters whose value is used to control the learning process) need to be adjusted as
they play an important role in the ﬁnal quality of the solution.

i=0

DNN are powerful enough to derive accurate non-linear features from the
training data, but these features are not robust. Indeed, adding a small amount
of noise at the input can cause these features to deviate and confuse the model.
In other words, the DNN is a very unbiased classiﬁer, but has a high variance.
Diﬀerent blocks can be used to implement these complex models. However,
in this paper, we will be using four types of blocks: the linear neural network, the
one-dimensional convolutional neural network, the activation functions (ReLU
and sigmoid) and the batch normalization.

Linear neural network. Linear neural networks applies a linear transforma-
tion to the incoming data: out = in.AT + b. Here we have θ = (A, b). The linear
neural network is also commonly named perceptron layer or dense layer.

One-dimensional convolutional neural network. The 1D-CNN applies a
convolution over a ﬁxed (multi-)temporal input signal. The 1D-CNN operation
can be seen as multiple linear neural networks (one per ﬁlter) where each one is
applied to a sub-part of the input. This sub-part is sliding, its size is kernel size,
its pitch is the stride and its start and end points depend on the padding.

Activation functions. The three activation functions that we discuss here
are the Rectiﬁed Linear Unit (ReLU), deﬁned as ReLU(x) = max(0, x), the
1+exp(−x) and the Heaviside step
sigmoid, deﬁned as Sigmoid(x) = σ(x) =
function, deﬁned as H(x) = 1
. This block, added between each layer of
the DNN, introduces the non-linear part of the model.

2 + sgn(x)

Batch normalization. Training samples are typically randomly collected in
batches to speed up the training process. It is thus usual to normalize the overall
tensor according to the batch dimension.

3 A First Look at Gohr’s CRYPTO 2019 Results

Since its release, the lightweight block cipher SPECK attracted a lot of external
cryptanalysis, together with its sibling SIMON (this was ampliﬁed by the fact

that no cryptanalysis was reported in the original speciﬁcations document [2]).
Many diﬀerent aspects of SPECK have been covered by these eﬀorts, but the
works from Dinur [6] and Song et al. [24] are the most successful advances on its
diﬀerential cryptanalysis aspect so far. Dinur [6] studied all versions of SPECK,
improving the best known diﬀerential characteristics (from [1, 3]) as well as de-
scribing a new key recovery strategy for this cipher. In particular, he devised a
4-round attack for 11 rounds of SPECK-32/64 using a 7 round diﬀerential char-
acteristic, that has a time complexity of 246 and data complexity of 222 (chosen
plaintexts).

Later, at CRYPTO’19, Gohr published a cryptanalysis work on SPECK-32/64
that is based on deep learning [11]. Gohr proposed a key-recovery attack on 11-
round SPECK-32/64 with estimated time complexity 238, improving the previous
best attack [6] in 246, albeit with a slightly higher data complexity: 214.5 cipher-
text pairs required. In this section, we will brieﬂy review Gohr’s results [11].

Overview. In his article, Gohr proposes multiple diﬀerential cryptanalysis of
SPECK, focusing on the input diﬀerence ∆in = 0x0040/0000. In this setting,
the aim is to distinguish real pairs, i.e., encryptions of plaintext pairs P, P (cid:48) such
that P ⊕ P (cid:48) = ∆in, from random pairs, which are the encryptions of random
pairs of plaintext with no ﬁxed input diﬀerence. Gohr compares a traditional
(pure) diﬀerential distinguisher with a distinguisher based on a DNN for 5 to 8
rounds of SPECK-32/64 and showed that the DNN performs better.

Pure diﬀerential distinguishers. Gohr computed the full DDT for the input
diﬀerence ∆in, using the Markov assumption. Then, to classify a ciphertext pair
(C, C(cid:48)), the probability p of the output diﬀerence C ⊕ C(cid:48) is read from the DDT
and compared to the uniform probability. Let ∆out = C ⊕ C(cid:48), then
if DDT (∆in → ∆out) > 1
232−1

(cid:40)

Classiﬁcation =

Real
Random otherwise

These distinguishers for reduced-round SPECK-32/64 are denoted Dnr, where
nr ∈ {5, 6, 7, 8} represents the number of rounds. The neural distinguishers are
denoted as Nnr.

Gohr’s neural distinguisher. We provide in Figure 2 a representation of
Gohr’s neural distinguisher. It is a deep neural network, whose main components
are:

1. Block 1: a 1D-CNN with kernel size of 1, a batch normalization and a ReLU

activation function

2. Blocks 2-i: one to ten layers with each layer consisting of two 1D-CNN with
kernel size of 3, each followed by batch normalization and a ReLU activation
function.

3. Block 3: a non-linear ﬁnal classiﬁcation block, composed of three percep-
tron layers separated by two batch normalization and ReLU functions, and
ﬁnished with a sigmoid function.

...

Fig. 2: The whole pipeline of Gohr’s deep neural network. Block 1 refers to the
initial convolution block, Block 2-1 to 2-10 refer to the residual block and Block
3 refers to the classiﬁcation block.

The input to the initial convolution block (Block 1) is a 4 × 16 matrix,
where each row corresponds to each 16-bit value in this order: Cl, Cr, C(cid:48)
l, C(cid:48)
r,
a convolution layer with 32 ﬁlters is then applied. The kernel size of this 1D-
CNN is 1, thus, it maps (Cl, Cr, C(cid:48)
r) to (f ilter1, f ilter2, ..., f ilter32). Each
f ilter is a non-linear combination of the features (Cl, Cr, C(cid:48)
r) after the ReLU
activation function depending on the value of the inputs and weights learned by
the 1D-CNN. The output of the ﬁrst block is connected to the input and added
to the output of the subsequent layer in the residual block (see Figure 3).

l, C(cid:48)

In the residual blocks (Blocks 2-i), both 1D-CNNs have a kernel of size 3,
a padding of size 1 and a stride of size 1 which make the temporal dimension
invariant across layers. At the end of each layer, the output is connected to the
input and added to the output of the subsequent layer to prevent the relevant
input signal from being wiped out across layers. The output of a residual block
is a (32 × 16) feature tensor (see Figure 4).

...

Fig. 3: Initial convolution block
(Block 1).

Fig. 4: The residual block (Blocks 2-i).

The ﬁnal classiﬁcation block takes as input the ﬂattened output tensor of
the residual block. This 512 × 1 vector is passed into three perceptron layers
(Multi-Layer Perceptron or MLP) with batch normalization and ReLU activation

functions for the ﬁrst two layers and a ﬁnal sigmoid activation function performs
the binary classiﬁcation (see Figure 5).

...

Fig. 5: The classiﬁcation block (Block 3).

Accuracy and eﬃciency of the neural distinguishers. For each pair, the
neural distinguishers outputs a real-valued score between 0 and 1. If this score
is greater than or equal to 0.5, the sample is classiﬁed as a real pair, and as a
random pair otherwise. The results given by Gohr are presented in Table 1. Note
that N7 and N8 are trained using some sophisticated methods (we refer to [11]
for more details on the training). We remark that Gohr’s neural distinguisher
has about 100,000 ﬂoating parameters, which is size eﬃcient considering the
accuracies obtained.

Table 1: Accuracies of neural distinguishers for 5, 6, 7 and 8 rounds (taken from
Table 2 of [11]). TPR and TNR denote true positive and true negative rates
respectively.

...

Real diﬀerences experiment. The neural distinguishers performed better
than the distinguishers using the full DDT, indicating that the neural distin-
guishers may learn something more than pure diﬀerential cryptanalysis. Gohr
explores this eﬀect with the real diﬀerences experiment. In this experiment,
instead of distinguishing a real pair from a random pair, the challenge is to
distinguish real pairs from masked real pairs, computed as (C ⊕ M, C(cid:48) ⊕ M ),
where M is a random 32-bit value. These experiments use the Nnr distin-
guishers directly, without retraining them for this new task. Table 2 shows the
accuracies of these distinguishers. Notice that this operation does not aﬀect

∆out = C ⊕ C(cid:48) = (C ⊕ M ) ⊕ (C(cid:48) ⊕ M ) and thus the output diﬀerence distri-
bution. However, the neural distinguishers are still able to distinguish real pairs
from masked pairs even without re-training for this particular purpose, which
shows that they do not just rely on the diﬀerence distribution.

Table 2: Accuracies of various neural distinguishers in the real diﬀerences exper-
iment.

...

Interpretation of Gohr’s Neural Network: a
Cryptanalysis Perspective

Interpretability of neural networks remains a highly researched area in machine
learning, but the focus has always been on improving the model and computa-
tional eﬃciency. We will discuss more about the interpretability in a machine
learning sense in Section 5. In this section, we want to ﬁnd out why and how
the neural distinguishers work in a cryptanalysis sense. In essence, we want to
answer the following question:

What type of cryptanalysis is Gohr’s neural distinguisher learning?

If the neural distinguisher is learning some currently-unknown form of cryptanal-
ysis, then we would like to extrapolate the additional statistics that it exploits.
If not, then we want to ﬁnd out what causes Gohr’s neural distinguishers to
perform better than pure diﬀerential attacks, and even improve state-of-the-art
attacks. With this question in mind, we perform a series of experiments and
analyses in order to come up with a reasonable guess, later validated by the
creation of a pure cryptanalysis-based distinguisher that matches the accuracy
of Gohr’s one.

Gohr’s neural distinguishers are able to correctly identify approximately
90.4%, 68.0% and 54.3% of the real ciphertext pairs (given by the true posi-
tive rates) for 5, 6 and 7 rounds of SPECK-32/64 respectively (see Table 1). We
will try to ﬁnd out what these ciphertext pairs are if there are any common
patterns and see whether we are able to identify and isolate them.

Choice of input diﬀerence. As a start, we looked into Gohr’s choice of input
diﬀerence: 0x0040/0000. This diﬀerence is part of a 9-round diﬀerential charac-
teristics from Table 7 of [1]. The reason given by Gohr is that this diﬀerence de-
terministically transits to a diﬀerence with low Hamming weight after one round.
Using constraint programming and techniques similar to [10], we found that the

best diﬀerential characteristics with a ﬁxed input diﬀerence of 0x0040/0000 for
5 rounds is 0x0040/0000 → 0x802a/d4a8, with probability of 2−13. In contrast,
when we do not restrict the input diﬀerence, the best diﬀerential characteristics
for 5 rounds is 0x2800/0010 → 0x850a/9520, with probability of 2−9. However,
when we trained the neural distinguishers to recognize ciphertext pairs with the
input diﬀerence of 0x2800/0010, the neural distinguishers performed worse (an
accuracy of 75.85% for 5 rounds). This is surprising as it is generally natural for
a cryptanalyst to maximize the diﬀerential probability when choosing a diﬀer-
ential characteristic. We believe this is explained by the fact that 0x0040/0000
is the input diﬀerence maximizing the diﬀerential probability for 3 or 4 rounds
of SPECK-32/64 (veriﬁed with constraint programming), which has the most
chances to provide a biased distribution one or two rounds later. Generally, we
believe that when using such neural distinguisher, a good method to choose
an input diﬀerence is to simply use the input diﬀerence leading to the highest
diﬀerential probability for nr − 1 or nr − 2 rounds.

Changing the inputs to the neural network. Gohr’s neural distinguishers
are trained using the actual ciphertext pairs (C, C(cid:48)) whereas the pure diﬀerential
distinguishers are only using the diﬀerence between the two ciphertexts C ⊕ C(cid:48).
Thus, it is unfair to compare both as they are not exploiting the same amount of
information. To have a fair comparison of the capability of neural distinguishers
and pure diﬀerential distinguishers, we trained new neural distinguishers using
C ⊕ C(cid:48), instead of (C, C(cid:48)). The results are an accuracy of 90.6% for 5 rounds,
75.4% for 6 rounds and 58.3% for 7 rounds. This shows us that when the neural
distinguishers are restricted to only have access to the diﬀerence distribution,
4 as
they do not perform as well as their respective Nnr, and similarly to Dnr
can be seen in Table 1. Therefore, this is another conﬁrmation (on top of the
real diﬀerences experiment conducted in [11]), that Gohr’s neural distinguishers
are learning more than just the distribution of the diﬀerences on the ciphertext.
With that information, we therefore naturally looked beyond just the diﬀerence
distribution at round nr.

4.1 Analyzing Ciphertext Pairs

In this section, we limit and focus the discussions and results mostly to 5 rounds
of SPECK-32/64. We recall that the last layer of the neural distinguisher is a
sigmoid activation function. Thus, its output is a value between 0 and 1. When
the score is 0.5 or more, the neural distinguisher predicts it as a real pair or
otherwise, random pair.

The closer a score is to 0.5, the least certain the neural distinguisher is on the
classiﬁcation. In order to know what are the traits that the neural distinguisher is
looking for, we segregate the ciphertext pairs that yield extreme scores, i.e. scores
that are either less than 0.1 (bad score) or more than 0.9 (good score). For the

4 Note that the new neural distinguishers are trained with 107 pairs, the same number

as in [11]

rest of this section, we label the ciphertext pairs as “bad” and “good” ciphertext
pairs and refer to the sets as B and G respectively. As we were experimenting
with them, we kept the keys (unique to each pair) that are used to generate the
ciphertext pairs. The goal now is to ﬁnd similarities and diﬀerences in these two
groups separately.

As we believe that most of the features the neural distinguishers learned is
diﬀerential in nature, we focus on the diﬀerentials of these ciphertext pairs. To
start, we did the following experiment (Experiment A):

1. Using 105 real 5-round SPECK-32/64 ciphertext pairs, extract the set G.
2. Obtain the diﬀerences of the ciphertext pairs and sort them by frequency
3. For each of the diﬀerences δ:

(a) Generate 104 random 32-bit numbers and apply the diﬀerence, δ to get

104 diﬀerent ciphertext pairs.

(b) Feed the pairs to the neural distinguisher N5 to obtain the scores.
(c) Note down the number of pairs that yield a score ≥ 0.5
In Table 3, we show the top 25 diﬀerences for 5 rounds of SPECK-32/64
with their respective score from the above experiment. Out of the ﬁrst 1000
diﬀerences, each records about 75% of the pairs scoring more than 0.5. Also,
there exist multiple pairs of diﬀerences such that one is more probable than the
other, and yet, it has a lower number of pairs classifying as real (e.g. No. 21
in Table 3). Thus, there is little evidence showing that if a diﬀerence is more
probable, then the neural distinguisher is necessarily more likely to recognize it.

Table 3: The top 25 diﬀerences (5 rounds of SPECK-32/64) in G with their
respective results for Experiment A as a percentage of how many pairs having a
score of ≥ 0.5 out of 104 pairs. Cnt refers to the number of diﬀerences obtained
in G.

...

Since the neural distinguishers outperform the ones with just the XOR input,
we started to look beyond just the diﬀerences at 5 rounds. We decided to partially

decrypt the ciphertext pairs from G for a few rounds and re-run Experiment A on
these partially decrypted pairs: for each pair, we compute the diﬀerence and for
each diﬀerence, we created 104 random plaintext pairs with these diﬀerences and
encrypted them to round nr using random keys. The results are very intriguing,
as compared to that of Table 3: almost all of the (top 1000) unique diﬀerences
obtained in this experiment achieved 99% or 100% of ciphertext pairs having a
score of ≥ 0.5.

We can see that the diﬀerences at rounds 3 and 4 (after decrypting 2 and 1
round respectively) start to show some strong biases. In fact, for all of the top
1000 diﬀerences at rounds 3 and 4, all 104 pairs × 1000 diﬀerences returned a
score of ≥ 0.55. With that, we conduct yet another experiment (Experiment B):

1. For all the ciphertext pairs in G, decrypt i rounds with their respective keys
and compute the corresponding diﬀerence. Denote the set of diﬀerences as
Diff5-i.

2. Generate 105 plaintext pairs with a diﬀerence of 0x0040/0000 with random

keys, encrypt to 4 rounds

3. If the pair’s diﬀerence is in Diff5-i, keep the pair. Otherwise, discard.
4. Encrypt the remaining pairs to 5 rounds and evaluate them using N5.

When i = 2, we obtain 1669 unique diﬀerences with a dataset size of 89,969.
97.86% of these ciphertext pairs yielded a score ≥ 0.5 (i.e. by this method, we
can isolate 88.04% of the true positive ciphertexts pair). Using i = 1, we have
128,039 unique diﬀerences and the size of the dataset is 74,077. While we could
get a cleaner set with 99.98% of these ciphertext pairs obtaining a score of ≥ 0.5,
we only managed to isolate 74.06% of the true positive pairs. Comparing with
the true positive rate of N5 from Table 1, which is 0.904 ± 8.33 × 10−4, the case
when i = 2 seems to be closer.
We also looked into the bias of the diﬀerence bits (the jth diﬀerence bit
refers to the jth bit index of C5−2 ⊕ C(cid:48)
5−2 where Cnr−i refers to the nr round
ciphertext decrypted by i rounds. Table 4 shows the diﬀerence bit biases of the
ﬁrst 1000 (most common) unique diﬀerences of ciphertext pairs in G and B
after decrypting two rounds. We assume that the neural distinguisher is able to
identify some bits at these rounds because they are signiﬁcantly more biased,
though both the set B and G are from the real distribution.

Now, we state the assumption required for our conjecture, which we will

verify experimentally in Section 4.3.

Assumption 1 Given a 5-round SPECK-32/64 ciphertext pair, N5 is able to
determine the diﬀerence of certain bits at rounds 3 and 4 with high accuracy.

Conjecture 1. Given a 5-round SPECK-32/64 ciphertext pair, N5 ﬁnds the dif-
ference of certain bits at round 3 and decides if the ciphertext pair is real or
random.

Interestingly, the diﬀerence bit biases after decrypting 1 and 2 rounds are
very similar (in their positions). We will provide an explanation in Section 4.2.

5 the diﬀerences were obtained experimentally.

Table 4: Diﬀerence bit bias of ciphertext pairs in G and B after decrypting 2
rounds. A negative (resp. positive) value indicates a bias towards ‘0’ (resp. ‘1’).
bit position 31

...
0.476 -0.454 -0.142 -0.006 0.025 0.084 -0.009 0.487 -0.473 -0.426 0.165 0.094 -0.006 0.019 -0.500 -0.500
0.031 -0.009 -0.015 -0.007 -0.014 -0.024 0.025 0.026 0.034 -0.005 -0.018 -0.021 0.006 0.009 0.079 -0.065

The exact truncated diﬀerentials are (∗ denotes no speciﬁc constraint, while 0
or 1 denotes the expected bit diﬀerence):

3 rounds: 10 ∗ ∗ ∗ ∗ ∗ 00 ∗ ∗ ∗ ∗ ∗ 00 10 ∗ ∗ ∗ ∗ ∗ 00 ∗ ∗ ∗ ∗ ∗ 10
4 rounds: 10 ∗ ∗ ∗ ∗ ∗ 10 ∗ ∗ ∗ ∗ ∗ 10 10 ∗ ∗ ∗ ∗ ∗ 10 ∗ ∗ ∗ ∗ ∗ 00

We refer to these particular truncated diﬀerential masks as T D3 and T D4 for
the following discussion. Using constraint programming, we evaluate that the
probabilities for these truncated diﬀerentials are 87.86% and 49.87% respectively.
In order to verify how much the neural distinguisher is relying on these bits, we
perform the following experiment (Experiment C):

1. Generate 106 plaintext pairs with initial diﬀerence 0x0040/0000 and 106
2. Encrypt all 106 plaintext pairs to 5 − i rounds. If a plaintext pair satisﬁes

random keys.

the T D5−i, then we keep it. Otherwise, it will be discarded.

3. Encrypt the remaining pairs to 5 rounds and evaluate them using N5.

Table 5: Results of Experiment C with T D3 and T D4. Proport. refers to the
number of true positive ciphertext pairs captured by the experiment.

...

Table 5 shows the statistics of the above experiment with 5 rounds of SPECK-
32/64. The true positive rates for ciphertext pairs that follow these are closer
to that of Gohr’s neural distinguisher. Now, there remains about 3% of the
ciphertext pairs yet to be explained (comparing the results of T D5−2 with N5).
The important point to note here is that the pairs we have identiﬁed are exactly
the ones veriﬁed by the neural distinguisher as well, by the nature of these
experiments. In other words, we managed to ﬁnd what the neural distinguisher
is looking for and not just another distinguisher that would achieve a good
accuracy by identifying a diﬀerent set of ciphertext pairs.

4.2 Deriving T D3 and T D4

With an input diﬀerence of 0x0040/0000, which has a deterministic transition to
0x8000/8000 in round 1, the diﬀerence will only start to spread after round 1 due

to the modular addition in the SPECK-32/64 round function. The inputs to the
modular addition at round 2 are 0x0100 and 0x8000 (cf. Figure 6). While there
are two active bits, only one of them will propagate the carry (as the other is
the MSB), resulting in multiple diﬀerences. Assuming a uniform distribution, the
carry has a probability 1
2 of propagating to the left. This causes the probability
of the various diﬀerentials to reduce by 1
2 as the carry bit propagates until b31
(bit position 31) is reached and any further carry will be removed by the modular
addition.

Fig. 6: The distribution of the possible output diﬀerences after passing through
the modular addition operation.

...

In Figure 7 and Figure 8, we show how the bits evolve along the most probable
diﬀerential path from round 1 (0x8000/8000) to round 4 (0x850a/9520). As
it passes through the modular addition operation, we highlight the bits that
have a relatively higher probability of being diﬀerent from the most probable
diﬀerential. The darker the color, the higher the probability of the diﬀerence
being toggled.

Figure 7 and Figure 8 show us why T D3 is important at round 3, and how
the active bits shift in SPECK-32/64 when we start with the input diﬀerence of
0x0040/0000. In every round, b31, (the leftmost bit) has a high probability of
staying active. This bit is then rotated to b24 before it goes into the modular
2 chance of switching from 1 → 0 or
addition operation. In each round, b26 has a 1
the other way round. b27 and b28 have a 1
8 chance respectively of switching.
This makes them highly volatile and therefore, unreliable. On the other hand,
the right part of SPECK-32/64 rotates by 2 to the left at the end of each round.
Because of the high rotation value in the left part of SPECK-32/64, low rotation
value of the right part of SPECK-32/64, and the fact that the left part is added
into the right part after the rotation, it takes about 3 to 4 rounds for the volatile
and unreliable bits to spread.

...

Fig. 7: The left (resp. right) part shows how the active bit from diﬀer-
ence 0x8000/8000 (resp. 0x8100/8102) propagates to diﬀerence 0x8100/8102
(0x8000/820a). The darker the color, the higher the probability (≥ 1
4 ) that it
has a carry propagated to.

...

Fig. 8: Showing how the active bit from diﬀerence 0x8000/820a propagates to
diﬀerence 0x850a/9520. The darker the color, the higher the probability (≥ 1
4 )
that it has a carry propagated to.

4.3 Verifying Assumption 1

To verify if Gohr’s neural distinguisher is able to recognize the truncated diﬀer-
ential, we retrain the neural distinguisher with a slight diﬀerence (Experiment
D):

1. Generate 107 plaintext pairs such that about 1

2 of the pairs satisfy T D3

(these are the positive pairs)

2. Encrypt the plaintext pairs for two rounds
3. Train the neural network to distinguish the two distributions, and validate
with the same hyper-parameters as in [11], with a depth of 1 in the residual
block.

After retraining, the neural distinguisher has an accuracy of 96.57% (TPR:
99.95%, TNR: 93.19%) This shows that the neural distinguisher has the ca-
pabilities to actually recognize the truncated diﬀerential with an outstanding
accuracy.

4.4 SPECK-32/64 Reduced to 6 Rounds

We perform Experiments C for 6 rounds of SPECK-32/64 as well. Table 6 shows
the comparison of the true positive results of rounds 5 and 6. While the results
are not as obvious as for the case of 5 rounds, we can still observe a similar trend
for 6 rounds.

...

Table 6: Results of SPECK-32/64 reduced to six rounds for Experiment C. Pro-
portion refers to the number of true positive ciphertext pairs captured by the
experiment.

4.5 Average Key Rank Diﬀerential Distinguisher

Taking into consideration the observations we presented in this section, we intro-
duce a new average key rank distinguisher that is not based on machine learning
and almost matches the accuracy as Gohr’s neural network for 5, 6 and 7 rounds
of SPECK-32/64. Here are the key considerations used in our distinguisher:

– The training set of Gohr’s neural network consists of 107 ciphertext pairs.
Thus, we restrict our distinguisher to only use 107 ciphertext pairs as well.
– If we do an exhaustive key search for two rounds, the time complexity will
be extremely high. Instead, we may need to limit ourselves to only one round
to match the complexity of the neural distinguishers.
– If we know the diﬀerence at round i, the i − 1 round diﬀerence for the right
part is known as well, since ri−1 = (li ⊕ ri) ≫ 2

With those pointers in mind, we created a distinguisher that uses an ap-
proximated DDT (aDDT); that is, a truncated DDT that is experimentally
constructed based on n ciphertext pairs. In this distinguisher, we use n = 107 to
ensure that both our distinguisher and the neural distinguishers have the same
amount of information. The idea of the distinguisher is to decrypt the last round,
nr, using all possible subkey bits that are relevant to the bits we are interested
in. Then, we compute the average of the probabilities of all partial decryptions
for a given pair, read from aDDT (nr − 1), to get a score. If the score is greater
than that of the random distribution, the distinguisher will return 1 (Real) and
0 (Random) otherwise. The bits we are interested in can be represented as an
AND mask, that is, a mask that has ‘1’ in the bit positions that we want to
consider the bit and ‘0’ for those we want to ignore. The mask value we have
chosen is 0xff8f/ff8f rather than the expected 0xc183/c183 as we believe the
truncated diﬀerential they are detecting is at nr − 2 rounds. Thus, other than
the bits that are identiﬁed earlier in this section, we decided to include more bits
to improve the accuracy. With the look-up table to the aDDT, we do not just
only match the data complexity (of the oﬄine training) of the Gohr’s neural
distinguishers, but at the same time, include the correlations between bits as
well.

The pseudocodes for creating the aDDT and the average key rank distin-
guisher can be found in Algorithm 1 and Algorithm 2 given in Supplementary
Materials. We applied the distinguisher for 5, 6 and 7 rounds of SPECK-32/64
and the results are given in Table 7. It shows that our distinguisher closely
matches the accuracies of Gohr’s neural distinguishers.

Degree of closeness. We now study the similarity between our distinguishers
and Gohr’s neural distinguishers. In particular, we are interested in whether the
classiﬁcations of the ciphertext pairs are the same for both distinguishers. To
verify this, we gave a set of 105 5-round ciphertext pairs (approx. 50,000 from
real and random distribution each) to both our average key rank distinguisher
and N5, and measured how many times did they have the same output. The
results for nr = 5 are shown in Table 8. We can see that about 97.6% of the
ciphertext pairs tested have the same classiﬁcation in both distinguishers. For
nr = 6, we achieved 94.98% of the pairs with the same classiﬁcation.

Complexity comparison. In our average key rank distinguisher, for each pair,
we perform the partial decryption of two ciphertexts, and a table lookup in
aDDT. In the partial decryption, we enumerate the 212 keys aﬀecting the right-
most 13 bits of δlnr−1 covered by our mask. Therefore, the complexity of our
distinguisher is 213 one-round SPECK-32/64 decryptions, and 213 table lookups.
Comparing its complexity with Gohr’s distinguishers is not trivial, as the op-
erations involved are diﬀerent. Gohr evaluates the complexity of his neural key
recovery by their runtime and an estimation of the number of speck encryptions
that could be performed at the same time on a GPU implementation. We pro-
pose to use the number of ﬂoating point multiplications performed by the neural
network instead. Let I and O respectively denote the number of inputs and out-

puts to one layer. The computational cost of going through a dense layer is I · O
multiplications. For 1D-CNN with kernel size ks = 1, a null padding, a stride
equal to 1 and F ﬁlters, with input size (I, T ) the cost is computed as I · F · T
multiplications. With the same input but with kernel size ks = 3, a padding
equal to 1, the cost is I · ks · F · T Applying these formulas to Gohr’s neural
network, we obtain a total of 137280 ≈ 217.07 multiplications. Note that we do
not account for batch normalizations and additions, which are dominated by the
cost of the multiplications. Using this estimation, it seems that our distinguisher
is slightly better in terms of complexity.

Table 7: Accuracy of the average key
rank distinguisher with a mask value
of 0xff8f/ff8f.

...

Table 8: Closeness of the outputs of N5
and average key rank distinguisher.

...

4.6 Discussion

Even though Gohr trained a neural distinguisher with a ﬁxed input diﬀerence, it
is unfair to compare the accuracy of neural distinguisher to that of a pure diﬀer-
ential cryptanalysis (with the use of DDT), since there are alternative cryptanal-
ysis methods that the neural distinguisher may have learned. The experiments
performed indicate that while Gohr’s neural distinguishers did not rely much on
diﬀerence at the nr round, they rely strongly on the diﬀerences at round nr − 1
and even more strongly at round nr − 2. These results support the hypothesis
that the neural distinguisher may learn diﬀerential-linear cryptanalysis [13] in
the case of SPECK. While we did not present any attacks here, using the MILP
model shown in [9], we veriﬁed that there are indeed many linear relations with
large biases for 2 to 3 rounds.

Unlike traditional linear cryptanalysis, which usually use independent char-
acteristics or linear hull involving the same plaintext and ciphertext bits, a well-
trained neural network is able to learn and exploit several linear characteristics
while taking into account their dependencies and correlations.

We believe that neural networks ﬁnd the easiest way to achieve the best
accuracy. In the case of SPECK, it seems that diﬀerential-linear cryptanalysis
would be a good ﬁt since it requires less data and the truncated diﬀerential
has a very high probability. Thus, we think that neural networks have the abil-
ity to eﬃciently learn short but strong diﬀerential, linear or diﬀerential-linear
characteristics for small block ciphers for a small number of rounds.

4.7 Application to AES-2-2-4 [7]

We are also interested in the capabilities of the neural distinguishers on a
Substitution-Permutation Network (SPN) cipher. We chose a small scale variant

of AES from [7] with the parameters: r = 2, c = 2, e = 4. We chose this cipher
as it has a small state size, which could be exhaustively searched through. AES-
2-2-8 would be a good choice as it also has a state size of 32-bit, however, our
distinguishers are not able to learn anything signiﬁcant. We trained AES-2-2-4
with 215 pairs, starting with an input diﬀerence of (1, 0, 0, 1). This input dif-
ference was chosen such that only after two rounds, all S-boxes will be active.
We trained them for 3 rounds and obtained an accuracy of 61.0%. In contrast,
we use the same number of pairs, we trained an aDDT distinguisher and we
obtained an accuracy of 62.3%.

To show the possibilities of relying purely on diﬀerences, we perform an
experiment similar to Experiment A. With the trained neural distinguisher, we
exhaust all possible 16-bit diﬀerences and we generate 100 random pairs for
each diﬀerence. Next, we feed the pairs to the neural distinguisher and count
the number of pairs in each basket of score: [0.0 − 0.1), [0.1 − 0.2), ..., [0.9 − 1.0].
Our result shows that for each diﬀerential, the 100 random pairs form a cluster
about a center similar to a Gaussian distribution. These results seem to suggest
the nature of the neural distinguisher for AES-2-2-4 is one that relies fully on
diﬀerential: giving a conﬁdence interval based on just the diﬀerence.

Interpretation of Gohr’s Neural Network: a Machine
Learning Perspective

In this section, we are exploring the following practical question:

Can Gohr’s neural network be replaced by a strategy inspired by both

diﬀerential cryptanalysis and machine learning?

We will demonstrate here that this is possible. First of all, it should be em-
phasized that DNNs often outperform mathematical modeling or standard ma-
chine learning approaches in supervised data-driven settings, especially on high-
dimensional data. It seems to be the case because correlations found between
input and output pairs during DNN training lead to more relevant character-
istics than those found by experts. In other words, Gohr’s neural distinguisher
seems to be capable of ﬁnding a property P currently unknown by cryptana-
lysts. One may ask if we could experimentally approach this unknown property
P that encodes the neural distinguisher behavior, using both machine learning
and cryptanalysis expertise. With this question in mind, we propose our best
estimate with a focus on 5 and 6 SPECK-32/64 rounds where the DNN achieves
accuracies of 92.9% and 78.8% in a real/random distinction setting and where
the full DDT approach can achieve accuracies of 91.1% and 75.8%. In our best
setting, we reach accuracy values of 92.3% and 77.9%.

Section 3 discusses in detail how Gohr’s neural distinguisher is modeled in
three blocks. Our objective here is to replace each of these individual blocks by a
more interpretable one, coming either from machine learning or from the crypt-
analysts’ point of view. This work is thus the result of the collaboration between

two worlds addressing the open question of deep learning interpretability. In the
course of the study, we set forth and challenged four conjectures to estimate the
property P learned by the DNN as detailed below.

5.1 Four Conjectures

Conjectures 2 & 3 aim to uncover Block 3 behavior. Conjecture 4 tackles Block
1 while Conjecture 5 concerns Block 2-i.

The DNN can not be entirely replaced by another machine learning
model. Ensemble-based machine learning models such as random forests [4]
and gradient boosting decision trees [8] are accurate and easier to interpret than
DNNs [14]. Nevertheless, DNNs outperform ensemble-based machine learning
models for most tasks on high-dimensional data such as images. However, with
only 64 bits of input, we could legitimately wonder whether the DNN could be
replaced by another ensemble-based machine learning model. Despite our small
size problem, our experiments reveal that other models signiﬁcantly decrease the
accuracy.

Conjecture 2. Gohr’s neural network outperforms other non-neuronal network
machine learning models.

Experiment. To challenge this conjecture, we tested multiple machine learn-
ing models, such as Random Forest (RF), Light Gradient Boosting Machine
(LGBM), Multi-Layer Perceptron (MLP), Support Vector Machine (SVM) and
Linear Regression (LR). They all performed equally. For the rest of this paper,
we will only consider LGBM [12] as an alternative ensemble classiﬁer for DNN
and MLP. LGBM is an extension of Gradient Boosting Decision Tree (GBDT) [8]
and we ﬁxed our choice on it because it is accurate, interpretable and faster to
train than RF or GBDT. In support of our conjecture, we established that the
accuracy for the LGBM model is signiﬁcantly lower than the one of the DNN
when the inputs are (Cl, Cr, C(cid:48)

r), see third column of Table 9.

l, C(cid:48)

Table 9: A comparison of the neural distinguisher and LGBM model for 5 round,
for 106 samples generated of type (Cl, Cr, C(cid:48)

...

The ﬁnal MLP block is not essential. As described above, we can not
replace the entire DNN with another non-neuronal machine learning model that
is easier to interpret. However, we may be able to replace the last block (Block

3) of the neural distinguisher performing the ﬁnal classiﬁcation, by an ensemble
model.

Conjecture 3. The MLP block of Gohr’s neural network can be replaced by an-
other ensemble classiﬁer.

Experiment. We successfully exchanged the ﬁnal MLP block for a LGBM model.
The reasons for choosing LGBM as a non-linear classiﬁer were detailed in the
previous experiment paragraph. The ﬁrst attempt is a complete substitution of
Block 3, taking the 512-dimension output of Block 2-10 as input. In the fourth
column of Table 9, we observe that this experiment leads to much better results
than the one from Conjecture 2, and even better results than the classical DDT
method D5 (+0.39%). To further improve the accuracy, we implemented a partial
substitution, taking only the 64-dimension output of the ﬁrst layer of the MLP as
input. As can be seen in the ﬁfth column from Table 9, the accuracy with those
inputs is now much closer to the DNN accuracy. In both cases, the accuracy
is close to the neural distinguisher, supporting our conjecture. At this point, in
order to grasp the unknown property P, one needs to understand the feature
vector at the residuals’ output.

The linear transformation on the inputs. We saw in Section 3 that Block
1 performs a linear transformation on the input. By looking at the weights of the
DNN ﬁrst convolution, we observe that it contains many opposite values. This
indicates that the DNN is looking for diﬀerences between the input features.
Consequently, we propose the following conjecture.

Conjecture 4. The ﬁrst convolution layer of Gohr’s neural network transforms
the input (Cl, Cr, C(cid:48)
r) into (∆L, ∆V, V0, V1) and a linear combination of those
terms.

l, C(cid:48)

Experiments. As the inputs of the ﬁrst convolution are binary, we could formally
verify our conjecture. By forcing to one all non-zero values of the output of this
layer, we calculated the truth-table of the ﬁrst convolution. We thus obtained
the boolean expression of the ﬁrst layer for the 32 ﬁlters. We observed that eight
ﬁlters were empty and the remaining twenty-four ﬁlters were simple. The ﬁlter
expressions are provided in Table 14 in the Supplementary Materials.

However, one may argue that setting all non-zero values to one is an over-
simpliﬁed approach. Therefore, we replaced the ﬁrst ReLU activation function
by the Heaviside activation function, and then we retrained the DNN. Since the
Heaviside function binarizes the intermediate value (as in [28]), we can estab-
lish the formal expression of the ﬁrst layer of the retrained DNN. This second
DNN had the same accuracy as the ﬁrst one and almost the same ﬁlter boolean
expression.

Finally, we trained the same DNN with the following entries (∆L, ∆V, V0, V1).
Using the same method as before, we established the ﬁlters’ boolean expressions.
This time, we obtained twenty ﬁve null ﬁlters and seven non-null ﬁlters, with the

following expressions: ∆L, V0∧ V1, ∆L, ∆L, V0∧ V1, ∆L∧ ∆V , ∆L∧ ∆V . These
observations support conjecture 4. Therefore, we kept only (∆L, ∆V, V0, V1) as
inputs for our pipeline.

The masked output distribution table. With regards to the remaining
residual block replacement, our ﬁrst assumption is that the DNN calculates a
shape close to the DDT in that residual block. However, two major properties of
the neural distinguisher prevent us from assuming that it is a DDT in the classical
sense of the term. The ﬁrst property, as explained in Section 3, is that the neural
distinguisher does not only rely on the diﬀerence distribution to distinguish real
pairs as presented in Table 2. The second speciﬁcity is that the DNN has only
approximately 100,000 ﬂoating parameters to perform classiﬁcation, which can
be considered as size eﬃcient. Our second assumption is therefore that the DNN
is able to compress the distribution table. We introduce the following deﬁnitions.

Output Distribution Table (ODT). We propose to compute a distribution table
on the values (∆L, ∆V, V0, V1) directly, instead of doing so on the diﬀerence
of the ciphertext pair (Cl ⊕ C(cid:48)
r). We call this new table an Output
Distribution Table (ODT) and it can be seen as a generalization of the DDT.
The entries of the ODT are 64 bits, which is not tractable for 107 samples. Also,
the DNN has only 100,000 parameters. The DNN is therefore able to compress
the ODT.

l, Cr ⊕ C(cid:48)

Masked Output Distribution Table (M-ODT). A compressed ODT means that
the input is not 64 bits, but instead hw bits, where hw represents the Hamming
weight of the mask. Let us consider a mask M ∈ Mhw with Mhw the ensemble
of 64-bits masks with Hamming weight hw and M = (M1, M2, M3, M4), with
Mi a 16-bit mask. Compressing the ODT therefore means applying the M mask
to all inputs. In our case, with I = (∆L, ∆V, V0, V1), we get IM = (∆L ∧
M1, ∆V ∧ M2, V0 ∧ M3, V1 ∧ M4) = I ∧ M , before computing the ODT. By
calculating that way, the number of ODT entries per mask decreases. It becomes
a function that depends only on hw and on the bit positions in the masks. It
is therefore a more compact representation of the complete ODT. However, it
turns out that if we consider only one mask, we get only one value per sample to
perform the classiﬁcation: P (Real|IM ), while the DNN has a ﬁnal vector size of
512. We considered several masks. Thus, by deﬁning the ensemble RM ∈ Mhw,
the set of relevant masks of Mhw, we can calculate for a speciﬁc input I =
(∆L, ∆V, V0, V1) the probability P (Real|IM ),∀M ∈ RM . Then, we concatenate
all the probabilities into a feature vector of size m = |RM|. We get the feature
. We are now able

F for the input I: F =(cid:0) P (Real|IM 1) P (Real|IM 2) ··· P (Real|IMm)(cid:1)T

to propose the ﬁnal conjecture.

Conjecture 5. The neural distinguisher internal data processing of Block 2-i can
be approached by:

1. Computing a distribution table for input (∆L, ∆V, V0, V1).

2. Finding several relevant masks and applying them to the input in order to

compress the output distribution table.

We abbreviate M-ODT this Masked-Output Distribution Table. Thus, the fea-
ture vector of the DNN can be replaced by a vector where each value represents
the probability stored in the M-ODT for each mask.

This approach enables us to replace Block 2-i of the DNN. Though, we still

need to clarify how to get the RM ensemble.

Extracting masks. Based on local interpretation methods, we can extract these
masks from the DNN. Indeed, these methods consist of highlighting the most
important bits of the entries for classiﬁcation. Thus, by sorting the entries ac-
cording to their score and by applying these local interpretation methods, we
can obtain the relevant masks.

5.2 Approximating the Expression of the Property P

From our conjectures, we hypothesized that we can approximate the unknown
property P that encodes the neural distinguisher behavior by the following:
– Changing (C, C(cid:48)) into I = (∆L, ∆V, V0, V1).
– Changing the 512-feature vector of the DNN by the feature vector of prob-

abilities F =(cid:0) P (Real|IM 1) P (Real|IM 2) ··· P (Real|IMm)(cid:1)T

– Changing the ﬁnal MLP block by the ensemble machine learning model

LGBM.

These points stand respectively for Block 1, Block 2-i and Block 3.

5.3 Implementation

In this section and based on the veriﬁed conjectures, we are describing the step-
wise implementation of our method. We consider that we have a DNN formed
with 107 data of type (∆L, ∆V, V0, V1) for 5 and 6 rounds of SPECK-32/64. We
developed a three-step approach:

1. Extraction of the masks from the DNN with a ﬁrst dataset.
2. Construction of the M-ODT with a second dataset.
3. Training of the ﬁnal classiﬁer from the probabilities stored in the M-ODT

with a third dataset.

Mask extraction from the DNN. We ﬁrst ranked 104 real samples accord-
ing to DNN score, as described in Section 4.1, in order to estimate the masks
from these entries. We used multiple local interpretation methods: Integrated
Gradients [26], DeepLift [22], Gradient Shap [15], Saliency maps [23], Shapley
Value [5], and Occlusion [27]. These methods score each bit according to their

importance for the classiﬁcation. Following averaging by batch and by method,
there were two possible ways to move forward. We could either assign a Ham-
ming weight or else set a threshold above which all bits would be set to one.
After a wide range of experiments, we chose the ﬁrst option and set the Ham-
ming weight to sixteen and eighteen (which turned out to be the best values in
our testing). This approach allowed us to build the ensemble RM of the relevant
masks.

Implementation details. We used the captum library6 which brings together
multiple methods on local interpretation. The dataset is divided into batches
of size about 2,500 and grouped by scores. The categories we used were: scores
from 1 to 0.9 (about 2,000 samples), scores from 0.9 to 0.5 (about 500 samples),
scores from 1 to 0.8 (about 2,100 samples) and scores from 1 to 0.5 (about 2,500
samples). This way, one score per method could be derived for each bit of each
sample. We then proposed several methods to average these importance scores
by bit of category: the sum of absolute values, the median of absolute values
and the average of absolute values. Then, we took the sixteen and eighteen best
values and we obtained a mask. There is one mask per score, one per local inter-
pretation method and one per averaging method. On average, for 5,000 samples
we generate about 100 relevant masks. Finally, with the methods available in
scikit-learn [20], we ranked the features and so the masks according to their per-
formance. After multiple repetitions of mask generation and selection at every
time, we obtained 50 masks that are eﬀective: they are provided in Table 15
in the Supplementary Materials. The ﬁnal ensemble of masks is the addition of
those 50 eﬀective masks and the generated relevant masks.

Constructing the M-ODT. Once the ensemble RM of relevant masks is de-
termined, we compute the M-ODT. Algorithm D (in Supplementary Materials)
describes our construction method which is similar to that of the DDT. The in-
puts of the algorithm include a second dataset composed of n = 107 real samples
of type I = (∆L, ∆V, V0, V1), and the set of relevant masks RM . The output is
the M-ODT dictionary with the mask as ﬁrst key, the masked input as second
key, and P (Real|I ∧ M ) = P (Real|IM ) as value.
The M-ODT dictionary is constructed as follow: ﬁrst, for each mask M in
RM , we compute the corresponding masked-dataset DM which is simply the
operation IM = I ∧ M for all I in D. Secondly we compute a dictionary U with
key the element of DM and with value the occurrences number of that element
in DM . Then, we compute for all element IM in DM the probability:

P (Real|IM ) =

P (IM|Real)P (Real)

P (IM|Real)P (Real) + P (IM|Random)P (Random)

6 https://github.com/pytorch/captum

with P (Real) = P (Random) = 0.5, P (IM|Random) = 2−HW (M ), HW (M ) being
n × U [IM ]. Finally we update
the Hamming weight of M and P (IM|Real) = 1
M-ODT as follow: M-ODT[M ][IM ] = P (Real|IM ).

vector Fj =(cid:0) P (Real|Ij∧M 1) P (Real|Ij∧M 2) ··· P (Real|Ij∧M m)(cid:1)T

Training the classiﬁer on probabilities. Upon building the M-ODT, we
can start training the classiﬁer. Given a third dataset D = {(input0, y0)...
(inputn, yn)}, with inputj a sample of type (C, C(cid:48)), transformed into (∆L, ∆V,
V0, V1) and the label yj ∈ [0, 1], with n = 106, we ﬁrst compute the feature
for all inputs and for
m = |RM|. Next, we determined the optimal θ parameters for the gθ model
according to Equation 1, with L being the square loss. Here, the gθ classiﬁer is
Light Gradient Boosting Machine (LGBM) [12].

Implementation details. Feature vectors are standardized. Model hyper-parameters
ﬁne-tuning has been achieved by grid search. Results were obtained by cross-
validation on 20% of the train set and the test set had 105 samples. Finally,
results are obtained on the complete pipeline for three diﬀerent seeds, ﬁve times
for every seed.

5.4 Results

The M-ODT pipeline was implemented with numpy, scikit-learn [20] and pytorch
[19]. The project code can be found at this URL address7. Our work station is
constituted of a GPU Nvidia GeForce GTX 970 with 4043 MiB memory and
four Intel core i5-4460 processors clocked at 3.20GHz.

General results. Table 10 shows accuracies of the DDT, the DNN and our M-
ODT pipeline on 5 and 6-round reduced SPECK-32/64 for 1.1 × 107 generated
samples. When compared to DNN and DDT, our M-ODT pipeline reached an
intermediate performance right below DNN. The main diﬀerence is the true
positive rate which is higher in our pipeline (this can be explained by the fact
that our M-ODT preprocessing only considers real samples). All in all, our M-
ODT pipeline successfully models the property P.

Matching. Table 11 summarizes the results of the quantitative correspondence
studies for the prediction between the two models. We compared the DNN
trained on samples type (∆L, ∆V, V0, V1) to our M-ODT pipeline. On 5 rounds,
we obtained a rate of 97.5% identical predictions. In addition, 91.3% were both
identical and equal to the label. On 6 rounds, matching prediction reduces down
to 93.1%.

We thus demonstrated that our method advantageously approximates the
performance of the neural distinguisher. With an initial linear transformation

7 https://github.com/AnonymousSubmissionEuroCrypt2021/A-Deeper-Look-at-

Machine-Learning-Based-Cryptanalysis

Table 10: A comparison of Gohr’s neural network, the DDT and our M-
ODT pipeline accuracies for around 150 masks generated each time, with input
(∆L, ∆V, V0, V1), LGBM as classiﬁer and 1.1 × 107 samples generated in total.
TPR and TNR refers to true positive and true negative rate respectively.

...

on the inputs, computing a M-ODT for a set of masks extracted from the DNN
and then classifying the resulting feature vector with LGBM, we achieved an
eﬃcient yet more easily interpretable approach than Gohr distinguishers. In-
deed, DNN obscure features are simply approached in our pipeline by F =
. Finally, we interpret the performance of
the classiﬁer globally (i.e. retrieving the decision tree) and locally (i.e. deduc-
ing which feature played the greatest role in the classiﬁcation for each sample)
as in [14]. Those results are not displayed as they are beyond the scope of the
present work, but they can be found in the project code.

Table 11: A comparison of Gohr’s neural network predictions and our M-
ODT pipeline predictions for around 150 masks generated each time, with input
(∆L, ∆V, V0, V1), LGBM as classiﬁer and 1.1 × 107 samples generated in total.

...

5.5 Application to SIMON Cipher

In order to check whether our approach could be generalized to other crypto-
graphic primitives, we evaluated our M-ODT method on 8 rounds of SIMON-
32/64 block cipher. Implementing the same pipeline, we enjoyed a 82.2% ac-
curacy for the classiﬁcation, whereas the neural distinguisher achieves 83.4%
accuracy. In addition, the matching rate between the two models was up to

92.4%. The slight deterioration in the results of our pipeline for SIMON can be
explained by the lack of eﬃcient masks as introduced in Section 5.3 for SPECK.

5.6 Discussions

From the cryptanalysts’ standpoint, one important aspect of using the neural
distinguisher is to uncover the property P learned by the DNN. Unfortunately,
while being powerful and easy to use, Gohr’s neural network remains opaque.

Our main conjecture is that the 10-layer residual blocks, considered as the
core of the model, are acting as a compressed DDT applied on the whole input
space. We model our idea with a Masked Output Distribution Table (M-ODT).
The M-ODT can be seen as a distribution table applied on masked outputs,
in our case (∆L, ∆V, V0, V1), instead of only the diﬀerence (Cl ⊕ C(cid:48)
l, Cr ⊕ C(cid:48)
r).
By doing so, features are no longer abstract as in the neural distinguisher. In
our pipeline, each one of the features is a probability for the sample to be real
knowing the mask and the input. In the end, with our M-ODT pipeline, we
successfully obtained a model which has only −0.6% diﬀerence accuracy with the
DNN and a matching of 97.3% on 5 rounds of SPECK-32/64. Additional analysis
of our pipeline (e.g. masks independence, inputs inﬂuence, classiﬁers inﬂuence)
are available into the project code. To the best of our knowledge, this work is
the ﬁrst successful attempt to exhibit the underlying mechanism of the neural
distinguisher. However, we note that a minor limitation of our method is that it
still requires the DNN to extract the relevant masks during the preparation of the
distinguisher. Since it is only during preparation, this does not remove anything
with regards to the interpretability of the distinguisher. Future work will aim
at computing these masks without DNN. All in all, our ﬁndings represent an
opportunity to guide the development of a novel, easy-to-use and interpretable
cryptanalysis method.

Improved Training Models

While in the two previous sections we focused on understanding how the neural
distinguisher works, here we will explain how one can outperform Gohr’s results.
The main idea is to create batches of ciphertext inputs instead of pairs.

We refer to batch input of size B, a group of B ciphertexts that are con-
structed from the same key. Here, we can distinguish two ways to train and
evaluate the neural distinguisher pipeline with batch input. The straightfor-
ward one is to evaluate the neural distinguisher score for each element of the
batch and then to take the median of the results. The second is to consider
the whole batch as a single input for a neural distinguisher. In order to do so,
we used 2-dimensional CNN (2D-CNN) where the channel dimension is the fea-
tures (∆L, ∆V, V0, V1). We should point out that, for sake of comparability with
Gohr’s work, we maintained the product of the training set size by the batch
size to be equal to 107. Both batch size-based challenging methods yielded sim-
ilar accuracy values (see Table 12). Notably, in both cases, we enjoyed 100%
accuracy on 5 and 6 rounds with batch sizes 10 and 50 respectively.

Table 12: Study of the batch size methods on the accuracies with (∆L, ∆V , V0,
V1) as input for 5 and 6 rounds.

...

Considering these encouraging outcomes, we extended the method to 7 rounds.
As the 7-round training is more sophisticated and the two previous methods are
equivalent, we decided to only apply the ﬁrst method (the averaging one), be-
cause it requires to train only one neural distinguisher. Results given in Table 13
conﬁrm our previous ﬁndings: with a batch size of 100, we obtain 99.7% accuracy
on 7 rounds. This remarkable outcome demonstrates the major improvement of
our batch strategy over those from earlier Gohr’s work.

Table 13: Study of the averaging batch size method on the 7-round accuracies
with (∆L, ∆V , V0, V1) as input.
1

...

Conclusion

In this article, we proposed a thorough analysis of Gohr’s deep neural network
distinguishers of SPECK-32/64 from CRYPTO’19. By carefully studying the clas-
siﬁed sets, we managed to uncover that these distinguishers are not only basing
their decisions on the ciphertext pair diﬀerence, but also the internal state diﬀer-
ence in penultimate and antepenultimate rounds. We conﬁrmed our ﬁndings by
proposing pure cryptanalysis-based distinguishers on SPECK-32/64 that match
Gohr’s accuracy. Moreover, we also proposed a new simpliﬁed pipeline for Gohr’s
distinguishers, that could reach the same accuracy while allowing a complete in-
terpretability of the decision process. We ﬁnally gave possible directions to even
improve over Gohr’s accuracy.

Our results indicate that Gohr’s neural distinguishers are not really pro-
ducing novel cryptanalysis attacks, but more like optimizing the information
extraction with the low-data constraints. Many more distinguisher settings, ma-
chine learning pipelines, types of ciphers should be studied to have a better
understanding of what machine learning-based cryptanalysis might be capable
of. Yet, we foresee that such tools could become of interest for cryptanalysts and
designers to easily and generically pre-test a primitive for simple weaknesses.

Our work also opens interesting directions with regards to interpretability of
deep neural networks and we believe our simpliﬁed pipeline might lead to better
interpretability in other areas than cryptography.

Acknowledgements

The authors are grateful to the anonymous reviewers for their insightful com-
ments that improved the quality of the paper. The authors are supported by
the Temasek Laboratories NTU grant DSOCL17101. We would like to thank
Aron Gohr for pointing out that the diﬀerential characteristics mentioned in the
attacks of Dinur’s [6] have been extended by one free round, thus, our previous
suggestion of extending Dinur’s attack by one round is invalid.

References

1. Abed, F., List, E., Lucks, S., Wenzel, J.: Diﬀerential cryptanalysis of round-reduced
simon and speck. In: Fast Software Encryption - FSE 2014. LNCS, vol. 8540, pp.
525–545. Springer (2014)

2. Beaulieu, R., Shors, D., Smith, J., Treatman-Clark, S., Weeks, B., Wingers, L.: The
SIMON and SPECK families of lightweight block ciphers. IACR Cryptol. ePrint
Arch. 2013, 404 (2013), http://eprint.iacr.org/2013/404

3. Biryukov, A., Roy, A., Velichkov, V.: Diﬀerential analysis of block ciphers SIMON
and SPECK. In: Cid, C., Rechberger, C. (eds.) Fast Software Encryption - FSE
2014. LNCS, vol. 8540, pp. 546–570. Springer (2014)

4. Breiman, L.: Random forests. Machine learning 45(1), 5–32 (2001)
5. Castro, J., G´omez, D., Tejada, J.: Polynomial calculation of the shapley value

based on sampling. Computers & Operations Research 36(5), 1726–1730 (2009)

6. Dinur, I.: Improved diﬀerential cryptanalysis of round-reduced speck. In: Selected

Areas in Cryptography - SAC 2014. pp. 147–164 (2014)

7. Duan, X., Yue, C., Liu, H., Guo, H., Zhang, F.: Attitude tracking control of
small-scale unmanned helicopters using quaternion-based adaptive dynamic surface
control. IEEE Access 9, 10153–10165 (2021), https://doi.org/10.1109/ACCESS.
2020.3043363

8. Friedman, J.H.: Greedy function approximation: a gradient boosting machine. An-

nals of statistics pp. 1189–1232 (2001)

9. Fu, K., Wang, M., Guo, Y., Sun, S., Hu, L.: Milp-based automatic search algorithms
for diﬀ erential and linear trails for speck. IACR Cryptol. ePrint Arch. 2016, 407
(2016)

10. Gerault, D., Minier, M., Solnon, C.: Constraint programming models for chosen key
diﬀerential cryptanalysis. In: Principles and Practice of Constraint Programming.
pp. 584–601. Springer (2016)

11. Gohr, A.: Improving attacks on round-reduced speck32/64 using deep learning. In:
Advances in Cryptology - CRYPTO 2019. LNCS, vol. 11693, pp. 150–179. Springer
(2019)

12. Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., Liu, T.Y.:
Lightgbm: A highly eﬃcient gradient boosting decision tree. In: Advances in neural
information processing systems. pp. 3146–3154 (2017)

13. Langford, S.K., Hellman, M.E.: Diﬀerential-linear cryptanalysis. In: Advances in

Cryptology - CRYPTO ’94. LNCS, vol. 839, pp. 17–25. Springer (1994)

14. Lundberg, S.M., Erion, G., Chen, H., DeGrave, A., Prutkin, J.M., Nair, B., Katz,
R., Himmelfarb, J., Bansal, N., Lee, S.I.: Explainable ai for trees: From local ex-
planations to global understanding. arXiv preprint arXiv:1905.04610 (2019)

15. Lundberg, S.M., Lee, S.I.: A uniﬁed approach to interpreting model predictions.

In: Advances in neural information processing systems. pp. 4765–4774 (2017)

16. Maghrebi, H., Portigliatti, T., Prouﬀ, E.: Breaking cryptographic implementations
using deep learning techniques. In: Security, Privacy, and Applied Cryptography
Engineering - SPACE 2016. pp. 3–26 (2016)

17. Mouha, N., Preneel, B.: A proof that the ARX cipher salsa20 is secure against
diﬀerential cryptanalysis. IACR Cryptol. ePrint Arch. 2013, 328 (2013), http:
//eprint.iacr.org/2013/328

18. Mouha, N., Wang, Q., Gu, D., Preneel, B.: Diﬀerential and linear cryptanalysis
using mixed-integer linear programming. In: Information Security and Cryptology
- Inscrypt 2011. pp. 57–76 (2011)

19. Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen,
T., Lin, Z., Gimelshein, N., Antiga, L., et al.: Pytorch: An imperative style, high-
performance deep learning library. In: Advances in neural information processing
systems. pp. 8026–8037 (2019)

20. Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O.,
Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., et al.: Scikit-learn: Machine
learning in python. the Journal of machine Learning research 12, 2825–2830 (2011)
21. Rivest, R.L.: Cryptography and machine learning. In: Advances in Cryptology -

ASIACRYPT ’91. pp. 427–439 (1991)

22. Shrikumar, A., Greenside, P., Kundaje, A.: Learning important features through

propagating activation diﬀerences. arXiv preprint arXiv:1704.02685 (2017)

23. Simonyan, K., Vedaldi, A., Zisserman, A.: Deep inside convolutional net-
works: Visualising image classiﬁcation models and saliency maps. arXiv preprint
arXiv:1312.6034 (2013)

24. Song, L., Huang, Z., Yang, Q.: Automatic diﬀerential analysis of ARX block ci-
phers with application to SPECK and LEA. In: Information Security and Privacy
- ACISP 2016. pp. 379–394 (2016)

25. Sun, S., G´erault, D., Lafourcade, P., Yang, Q., Todo, Y., Qiao, K., Hu, L.: Analysis
of aes, skinny, and others with constraint programming. IACR Trans. Symmetric
Cryptol. 2017(1), 281–306 (2017)

26. Sundararajan, M., Taly, A., Yan, Q.: Axiomatic attribution for deep networks.

arXiv preprint arXiv:1703.01365 (2017)

27. Zeiler, M.D., Fergus, R.: Visualizing and understanding convolutional networks.

In: European conference on computer vision. pp. 818–833. Springer (2014)

28. Zhou, S., Wu, Y., Ni, Z., Zhou, X., Wen, H., Zou, Y.: Dorefa-net: Training
low bitwidth convolutional neural networks with low bitwidth gradients. CoRR
abs/1606.06160 (2016), http://arxiv.org/abs/1606.06160

Supplementary Materials

...