The Hashed Fractal Key Recovery (HFKR) Problem: From Symbolic Path Inversion to Post-Quantum Cryptographic Keys

Mohamed Aly Bouke

REVIEW 4 major objections 4 minor 17 references

The Hashed Fractal Key Recovery (HFKR) Problem: From Symbolic Path Inversion to Post-Quantum Cryptographic Keys

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proposes deriving post-quantum keys by hashing chaotic symbolic walks, replacing algebraic hardness assumptions with symbolic complexity plus hash diffusion.

desk verdict A named primitive that is really SPIP-plus-hashing, with the post-quantum security claim resting on an unproved reduction and a botched BLAKE3 comparison; the reproducible experiments don't rescue the central claim. read the letter →

arxiv 2506.04383 v1 pith:42GSIYW6 submitted 2025-06-04 cs.CR

Mohamed Aly Bouke This is my paper

classification cs.CR

keywords HashedFractalKeyRecoverySymbolicPathInversionProblempost-quantumcryptographydynamicshash-basedobfuscationdimensionchaostheorynon-algebraichardnessassumptions

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper is trying to establish that a cryptographic key can be generated by hashing a chaotic symbolic walk, with no factorization, discrete-logarithm, or lattice assumption underneath. It builds on the Symbolic Path Inversion Problem (SPIP), whose hardness it takes as given, and adds a hash layer: the concatenated walk is run through SHA3-512, SHAKE256, or BLAKE3 to obtain a fixed-size key. Over 250 perturbation trials, the walks show a box-counting dimension that stabilizes near 1.06, and SHA3-512 and SHAKE256 convert a small path change into about 255 flipped bits out of 512, which the paper interprets as hash-based amplification of symbolic divergence. If the scheme is sound, it would supply a lightweight, algebra-free route to key generation that is meant to hold up against quantum attackers.

What carries the argument

The mechanism is the symbolic trajectory $P = \{x_0, x_1, \ldots, x_n\}$ generated by the recursion $x_{i+1} = \lfloor A_i x_i + b_i + \delta_i \rfloor$, where $A_i$ is a contractive linear map, $b_i$ a bounded translation, and $\delta_i$ bounded noise; the path is concatenated into a message $M = x_0 \| x_1 \| \cdots \| x_n$ and hashed to a key $k = H(M)$. The trajectory's role is to supply combinatorial and fractal entropy, while $H$ acts as an entropy amplifier and one-way compressor. The box-counting dimension, which stabilizes near 1.06 in the experiments, is the paper's quantitative handle on the fractal character of the symbolic paths.

What would settle it

Enumerate all $m^n$ symbolic paths for a small trajectory length $n$, hash each one, and measure the min-entropy of the resulting key distribution; if it falls well below 256 bits, HFKR cannot provide the claimed security regardless of SPIP's hardness.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that symbolic chaos and hashing form a usable non-algebraic primitive. Contractive affine maps over $\mathbb{Z}^2$, perturbed by bounded noise and rounded down, generate paths whose combinatorial growth and fractal structure can be measured, and a cryptographic hash compresses those paths into keys while amplifying any change in the path. The author claims that this synergy is what gives HFKR its post-quantum character: SPIP supplies intractability, and the hash supplies one-wayness and the avalanche effect, so seeing only the key should not reveal the path. The empirical results are read as confirming the entropy foundation, with the box-counting dimension converging to roughly 1.06 and SHA3-512 and SHAKE256 reaching mean Hamming distances near 255 bits, while BLAKE3 shows weaker diffusion. The paper does not claim to prove the full reduction from SPIP hardness to hashed key recovery; that proof is explicitly left for future work.

Load-bearing premise

The scheme's security rests on SPIP being genuinely intractable for the short, stochastic walks used here and on the hash layer inheriting that hardness; the paper defers the proof of the inheritance step.

Editorial extensions

If this is right

If HFKR is sound for the proposed parameters, post-quantum key generation would no longer be tied to algebraic problems such as lattice inversion, opening a distinct design space for key generation.
Hash selection becomes part of the security argument: SHA3-512 and SHAKE256 preserve symbolic divergence well, while BLAKE3's weaker diffusion makes it a poor choice for this construction.
The stabilized fractal dimension near 1.06 gives a quantitative target for choosing trajectory length in future instantiations of the scheme.
The construction's simplicity supports the paper's claim that it can run in constrained environments, provided the missing SPIP-to-HFKR reduction is eventually supplied.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural stress test the paper leaves implicit is to enumerate all symbolic paths for small $n$ and compute the true min-entropy of $H(M)$; if that entropy falls below the key length, the advertised security bound would not hold.
Because the implementation samples fresh affine maps at every step as a stochastic extension of SPIP, a complete treatment would need a reduction showing that this extension preserves the hardness of the original SPIP problem.
The pairing of p-values near 1.0000 with BLAKE3's lower mean Hamming distance suggests that uniform bit-flip statistics and strong diffusion are separable properties; a minimum-Hamming-distance metric over many perturbations would make the distinction explicit.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

This paper introduces the "Hashed Fractal Key Recovery problem" — basically the author's earlier Symbolic Path Inversion Problem (SPIP) followed by a hash. The genuinely new part is the packaging: a key-generation algorithm built on symbolic walks with a hash on top, plus some empirical diffusion measurements. The central claim, that HFKR is a post-quantum-secure non-algebraic primitive, is not established anywhere in the manuscript. The paper itself says in Section 5 that reductions linking SPIP inversion to hash-based recovery are future work. That is not a cosmetic gap: an adversary wanting the key doesn't have to invert the symbolic path at all. SPIP hardness alone, even if true, doesn't transfer through a hash.

What the paper does well: it is clearly written, the algorithm is specified, the code is on GitHub, and the author openly flags the missing reductions. The avalanche experiments on SHA3-512 and SHAKE256 are standard but honestly reported. There is no sign of fabrication; the numbers are plausible.

The soft spots, in proportion:

1. The BLAKE3 comparison is invalid. BLAKE3's default output is 256 bits; SHA3-512 outputs 512 bits. The paper reports BLAKE3's mean Hamming distance of ~128 as "weaker diffusion," but 128/256 is exactly the expected avalanche rate. The output lengths are never disclosed, so the comparison is apples-to-oranges. That is a real, identifiable error.

2. No min-entropy argument for the hashed keys. Section 4 measures the difference between perturbed and unperturbed hashes (avalanche), not the distribution of keys produced by Algorithm 1 over the parameter and noise space. The fractal dimension stabilizing near 1.06 is close to a space-filling curve in 2D, but that says little about the entropy of the concatenated path; a low-dimensional trajectory can still have low complexity.

3. The chi-square p-values of 1.0000 are interpretative noise. The author half-acknowledges this in the text, but then uses "higher chi-square statistic" to claim SHA3/Shake diffuse better. That's not what the test shows.

Overall: this is a proposal with a named problem and some reproducible measurements, but the load-bearing security claim is an assertion, not a result. The paper doesn't deserve a serious referee on its current form. If the author later supplies a real reduction and a min-entropy bound, the underlying direction might be worth a second look.

Recommendation: desk reject. It is an honest, organized preprint, but the central claim is unsupported and the empirical comparison is flawed. Cite it only as an example of why avalanche tests alone don't establish key-recovery hardness.

Referee Report

4 major / 4 minor

Summary. The paper introduces the Hashed Fractal Key Recovery (HFKR) problem, a proposed non-algebraic cryptographic construction. Algorithm 1 generates a symbolic trajectory over Z^2 via noisy contractive affine maps, and Eq. (2) derives a cryptographic key as k = H(x0 || ... || xn). The paper claims that HFKR achieves post-quantum security from the 'synergy' of symbolic fractality and hash-based entropy amplification. The empirical component measures the fractal dimension of generated trajectories (stabilizing near 1.06) and compares the avalanche/diffusion properties of SHA3-512, SHAKE256, and BLAKE3 across 250 perturbation trials, reporting Hamming distances, bit-flip rates, entropy deviations, and chi-square uniformity tests.

Significance. If the central claim were established, a non-algebraic, structure-free key-recovery-hard primitive would be a notable contribution to post-quantum cryptography, especially given the scarcity of constructions that avoid lattice and number-theoretic assumptions. The paper has clear strengths: the construction is precisely specified, the empirical methodology is reproducible (the code is publicly linked), and the diffusion metrics are concrete and falsifiable. However, the significance is sharply limited because the paper does not provide a security model, a reduction from any established hard problem to key recovery, or an entropy bound for the produced keys. As it stands, the paper supplies empirical evidence about trajectory geometry and hash avalanche behavior, but not a cryptographic hardness result.

major comments (4)

[Section 5, Eq. (2)] The paper's central security claim is unsupported by any reduction. Section 5 explicitly defers 'formalizing hardness reductions that link SPI style symbolic inversion with hash-based obfuscation under standard adversarial models' to future work. This is not a cosmetic gap: an adversary seeking k = H(x0 || ... || xn) need not recover the symbolic path x0,...,xn at all, so the cited #P-hardness/PSPACE-hardness of SPIP [15] does not, by itself, imply key-recovery hardness. The manuscript needs either a reduction from key recovery to SPIP, or a direct lower-bound argument for the HFKR problem; neither appears.
[Section 4, Table 1 and Figure 2] The BLAKE3 comparison is invalid because the output lengths are not matched. BLAKE3's default output is 256 bits, while SHA3-512 and SHAKE256 outputs are 512 bits in this experiment. For a 256-bit random output, the expected Hamming distance between two independent samples is 128 bits; the reported mean of 128.21 for BLAKE3 is therefore exactly what ideal diffusion would produce, not 'weaker diffusion.' The conclusion that BLAKE3 exhibits 'notably weaker diffusion' is an artifact of comparing raw bit counts across different digest sizes. The comparison should use a normalized metric (e.g., Hamming distance divided by output length) or equal-length outputs.
[Section 4, entropy analysis] The entropy analysis does not measure the entropy of the key distribution. The reported 'entropy difference ΔH' is computed between perturbed and unperturbed hash outputs, which is an avalanche measure, not an estimate of the min-entropy of k over the parameter and noise space of Algorithm 1. The paper never bounds the min-entropy of the concatenated trajectory or of the resulting key, so there is no basis for claiming resistance to brute-force or guessing attacks. The observed fractal dimension stabilizing near 1.06 is, if anything, a warning sign: a nearly one-dimensional object has low effective dimensionality, which could imply low path entropy rather than high entropy.
[Section 3, Algorithm 1 and surrounding text] The paper calls HFKR a 'novel cryptographic hardness assumption' but never defines the assumption formally: there is no adversarial success experiment, no advantage function, no security parameter, and no statement of what is conjectured to be hard. In particular, the phrase 'post-quantum security' is used throughout without a quantum adversary model (e.g., QROM or quantum circuit model). As a result, the paper's abstract and conclusion go beyond what the evidence supports: the empirical results are consistent with standard hash-function avalanche properties and do not establish a new hardness assumption.

minor comments (4)

[References [16] and [17]] The citation to Barnsley [16] appears mismatched: reference [16] is listed as an arXiv preprint with no author name, while the text attributes the fractal-symbolic connection to Barnsley and Kitchens. Please correct the reference list or the in-text attribution.
[Section 4, captions of Figures 2-4] Figure captions do not specify the sample size, the perturbation offsets, or the units on the axes. The text mentions 250 trials and 5 perturbation offsets, but the captions should state this directly for readability.
[Section 5] There is a typographical error: 'SPI style symbolic inversion' should be 'SPIP-style symbolic inversion.'
[Section 3, Eq. (1)] The claim that the trajectory is 'non-invertible due to stochastic perturbations and rounding' is asserted without proof. Rounding is indeed non-injective, but the statement would benefit from a precise argument or a reference, especially since invertibility is not equivalent to hardness of inversion.

Circularity Check

0 steps flagged · score 0.0 of 10

No construction-level circularity: the HFKR security claim is not derived, but the gap is a missing reduction, not a circular one.

full rationale

The paper's empirical results (fractal dimension, Hamming distance, bit-flip rates, chi-square statistics) are self-contained measurements, not predictions obtained by fitting parameters to the data, and the construction k = H(x0 || x1 || ... || xn) in Eq. (2) is not defined in terms of the security claim it is supposed to support. The hardness premise is the author's separate SPIP result [15], which concerns symbolic path inversion rather than HFKR key recovery; citing it is a self-citation, but not a circular one, because SPIP hardness does not assume HFKR security and is a different target result. The genuine defect is that Section 5 explicitly defers 'formalizing hardness reductions that link SPI style symbolic inversion with hash-based obfuscation,' so the paper never proves SPIP hardness transfers to key recovery, nor does it bound the min-entropy of generated keys. That is an omitted proof and an unsupported security claim, not a reduction of the conclusion to its inputs by construction. Accordingly, no circular step can be exhibited under the required standard.

Assumptions & free parameters 5 free parameters · 3 assumptions · 1 invented entities

The central claim leans on three unpaid inputs: the self-cited SPIP hardness theorem, the assumption that cryptographic hashes faithfully amplify symbolic path entropy, and the unquantified min-entropy of the generated trajectories. The paper supplies no proof, reduction, or machine-checked certificate for any of these, and several experimental parameters are left unspecified.

free parameters (5)

Affine map parameters (a11, a12, a21, a22)
Algorithm 1 takes these as input, but the paper never gives concrete values or the distribution used; they determine the trajectory geometry and any entropy claims.
Translation bounds [bmin, bmax]
Input to Algorithm 1, unspecified; controls how far the walk can move each step.
Noise bound epsilon
Input to Algorithm 1, unspecified; perturbation size directly affects divergence measurements.
Trajectory length n = 128, 500, 2000, 5000
Chosen by hand for experiments; fractal dimension is reported as increasing with n and stabilizing near 1.06.
Hash output length = not stated (BLAKE3 likely 256-bit)
The Hamming-distance comparison assumes comparable output sizes, but BLAKE3's output length is not disclosed; this affects the weaker diffusion conclusion.

assumptions (3)

ad hoc to paper SPIP is #P-hard and PSPACE-hard (reference [15])
The paper relies on this self-cited result as the hardness foundation for HFKR, with no independent proof or machine-checked certificate visible.
domain assumption Hash functions act as faithful entropy amplifiers for structured symbolic paths
The paper treats the hash's avalanche property as sufficient to turn symbolic divergence into cryptographic security; this is asserted, not derived from a model.
domain assumption Symbolic path entropy is sufficient to avoid key-recovery attacks
No min-entropy lower bound or key-recovery model is given; the security claim assumes the path itself cannot be guessed or inverted.

invented entities (1)

HFKR hardness assumption
purpose: Postulate that recovering the hashed symbolic path or key is intractable without algebraic structure
No falsifiable handle is provided: no reduction, no concrete parameter set, no attack analysis. The assumption is asserted to follow from SPIP [15].

how reviews work

0 comments

Cite this review

Pith. "Pith review of The Hashed Fractal Key Recovery (HFKR) Problem: From Symbolic Path Inversion to Post-Quantum Cryptographic Keys." pith.science (2026). https://pith.science/paper/42GSIYW6

@misc{pith2026250604383,
  author       = {Pith},
  title        = {Pith review of: The Hashed Fractal Key Recovery (HFKR) Problem: From Symbolic Path Inversion to Post-Quantum Cryptographic Keys},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/42GSIYW6}},
  note         = {Machine review of arXiv:2506.04383}
}

abstract

Classical cryptographic systems rely heavily on structured algebraic problems, such as factorization, discrete logarithms, or lattice-based assumptions, which are increasingly vulnerable to quantum attacks and structural cryptanalysis. In response, this work introduces the Hashed Fractal Key Recovery (HFKR) problem, a non-algebraic cryptographic construction grounded in symbolic dynamics and chaotic perturbations. HFKR builds on the Symbolic Path Inversion Problem (SPIP), leveraging symbolic trajectories generated via contractive affine maps over $\mathbb{Z}^2$, and compressing them into fixed-length cryptographic keys using hash-based obfuscation. A key contribution of this paper is the empirical confirmation that these symbolic paths exhibit fractal behavior, quantified via box counting dimension, path geometry, and spatial density measures. The observed fractal dimension increases with trajectory length and stabilizes near 1.06, indicating symbolic self-similarity and space-filling complexity, both of which reinforce the entropy foundation of the scheme. Experimental results across 250 perturbation trials show that SHA3-512 and SHAKE256 amplify symbolic divergence effectively, achieving mean Hamming distances near 255, ideal bit-flip rates, and negligible entropy deviation. In contrast, BLAKE3 exhibits statistically uniform but weaker diffusion. These findings confirm that HFKR post-quantum security arises from the synergy between symbolic fractality and hash-based entropy amplification. The resulting construction offers a lightweight, structure-free foundation for secure key generation in adversarial settings without relying on algebraic hardness assumptions.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

17 extracted references · 15 canonical work pages

[15]

M. A. Bouke, On the intractability of chaotic symbolic walks: Toward a non-algebraic post-quantum hardness assumption, 2025. arXiv: 2505 . 22644 [cs.CR] . [Online]. Available: https://arxiv.org/abs/ 2505.22644

work page Pith review arXiv 2025
[1]

Rsa chips (past/present/future),

R. L. Rivest, “Rsa chips (past/present/future),” inWork- shop on the Theory and Application of of Cryptographic Techniques, Springer, 1984, pp. 159–165

work page 1984
[2]

A blockchain-integrated chaotic fractal en- cryption scheme for secure medical imaging in industrial iot settings,

S. Inam, S. Kanwal, M. Batool, S. Al-Otaibi, and M. M. Jamjoom, “A blockchain-integrated chaotic fractal en- cryption scheme for secure medical imaging in industrial iot settings,” Scientific Reports, vol. 15, no. 1, p. 7652, 2025

work page 2025
[3]

Recommendation for pair-wise key- establishment schemes using discrete logarithm cryptog- raphy,

E. Barker, L. Chen, S. Keller, A. Roginsky, A. Vas- silev, and R. Davis, “Recommendation for pair-wise key- establishment schemes using discrete logarithm cryptog- raphy,” National Institute of Standards and Technology, Tech. Rep., 2017

work page 2017
[4]

Morse and G

M. Morse and G. A. Hedlund, Symbolic dynamics. Insti- tute for Advanced Study, 1966

work page 1966
[5]

On the use of dynamical systems in cryptog- raphy,

S. Everett, “On the use of dynamical systems in cryptog- raphy,”Chaos, Solitons & Fractals, vol. 183, p. 114 952, 2024

work page 2024
[6]

Synchronization of chaotic and hyperchaotic nonlinear dynamical systems and their numerous applications: A review,

R. M. Bora and B. B. Sharma, “Synchronization of chaotic and hyperchaotic nonlinear dynamical systems and their numerous applications: A review,”Journal of Applied Nonlinear Dynamics, vol. 14, no. 02, pp. 313– 341, 2025

work page 2025
[7]

Post-quantum cryptosystems for internet-of- things: A survey on lattice-based algorithms,

R. Asif, “Post-quantum cryptosystems for internet-of- things: A survey on lattice-based algorithms,” IoT, vol. 2, no. 1, pp. 71–91, 2021

work page 2021

Show all 17 references

[8]

Post- quantum stealth address protocols,

M. Mikic, M. Srbakoski, and S. Praska, “Post- quantum stealth address protocols,” arXiv preprint arXiv:2501.13733, 2025

2025 arXiv
[9]

Algorithmic security is insufficient: A compre- hensive survey on implementation attacks haunting post- quantum security,

A. C. Canto, J. Kaur, M. M. Kermani, and R. Azarder- akhsh, “Algorithmic security is insufficient: A compre- hensive survey on implementation attacks haunting post- quantum security,” arXiv preprint arXiv:2305.13544 , 2023

2023 arXiv
[10]

Post-quantum lattice-based cryp- tography implementations: A survey,

H. Nejatollahi, N. Dutt, S. Ray, F. Regazzoni, I. Banerjee, and R. Cammarota, “Post-quantum lattice-based cryp- tography implementations: A survey,”ACM Computing Surveys (CSUR), vol. 51, no. 6, pp. 1–41, 2019

2019
[11]

M. A. Bouke, Fractal attractors in random nonlin- ear iterated function systems: Existence, stability, and dimensional properties , 2025. arXiv: 2505 . 18849 [math.DS]. [Online]. Available: https://arxiv. org/abs/2505.18849

2025 arXiv
[12]

Fractal-based hybrid cryptosys- tem: Enhancing image encryption with rsa, homomorphic encryption, and chaotic maps,

D. E. Mfungo and X. Fu, “Fractal-based hybrid cryptosys- tem: Enhancing image encryption with rsa, homomorphic encryption, and chaotic maps,” Entropy, vol. 25, no. 11, p. 1478, 2023

2023
[13]

A quantum- secure cryptographic algorithm integrating fractals and prime numbers,

G. Iovane, E. Benedetto, A. Di Lauro,et al., “A quantum- secure cryptographic algorithm integrating fractals and prime numbers,” Applied Sciences, vol. 14, pp. 10 138– 10 150, 2024

2024
[14]

Effi- ciency analysis for public key systems based on fractal functions,

N. M. AL-Saidi, M. R. Said, and A. M. Ahmed, “Effi- ciency analysis for public key systems based on fractal functions,” Journal of Computer Science, vol. 7, no. 4, p. 526, 2011

2011
[16]

Symbolic iterated function systems, fast basins and frac- tal manifolds,

“Symbolic iterated function systems, fast basins and frac- tal manifolds,” arXiv preprint arXiv:1308.3819,

arXiv
[17]

B. P. Kitchens, Symbolic dynamics: one-sided, two-sided and countable state Markov shifts. Springer Science & Business Media, 2012. Page 5 of 5

2012

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [15]

M. A. Bouke, On the intractability of chaotic symbolic walks: Toward a non-algebraic post-quantum hardness assumption, 2025. arXiv: 2505 . 22644 [cs.CR] . [Online]. Available: https://arxiv.org/abs/ 2505.22644

work page Pith review arXiv 2025

[2] [1]

Rsa chips (past/present/future),

R. L. Rivest, “Rsa chips (past/present/future),” inWork- shop on the Theory and Application of of Cryptographic Techniques, Springer, 1984, pp. 159–165

work page 1984

[3] [2]

A blockchain-integrated chaotic fractal en- cryption scheme for secure medical imaging in industrial iot settings,

S. Inam, S. Kanwal, M. Batool, S. Al-Otaibi, and M. M. Jamjoom, “A blockchain-integrated chaotic fractal en- cryption scheme for secure medical imaging in industrial iot settings,” Scientific Reports, vol. 15, no. 1, p. 7652, 2025

work page 2025

[4] [3]

Recommendation for pair-wise key- establishment schemes using discrete logarithm cryptog- raphy,

E. Barker, L. Chen, S. Keller, A. Roginsky, A. Vas- silev, and R. Davis, “Recommendation for pair-wise key- establishment schemes using discrete logarithm cryptog- raphy,” National Institute of Standards and Technology, Tech. Rep., 2017

work page 2017

[5] [4]

Morse and G

M. Morse and G. A. Hedlund, Symbolic dynamics. Insti- tute for Advanced Study, 1966

work page 1966

[6] [5]

On the use of dynamical systems in cryptog- raphy,

S. Everett, “On the use of dynamical systems in cryptog- raphy,”Chaos, Solitons & Fractals, vol. 183, p. 114 952, 2024

work page 2024

[7] [6]

Synchronization of chaotic and hyperchaotic nonlinear dynamical systems and their numerous applications: A review,

R. M. Bora and B. B. Sharma, “Synchronization of chaotic and hyperchaotic nonlinear dynamical systems and their numerous applications: A review,”Journal of Applied Nonlinear Dynamics, vol. 14, no. 02, pp. 313– 341, 2025

work page 2025

[8] [7]

Post-quantum cryptosystems for internet-of- things: A survey on lattice-based algorithms,

R. Asif, “Post-quantum cryptosystems for internet-of- things: A survey on lattice-based algorithms,” IoT, vol. 2, no. 1, pp. 71–91, 2021

work page 2021