Sirvi Autor "Helm, Louie" järgi

Nüüd näidatakse 1 - 3 3

listelement.badge.access-status Avatud juurdepääs ,
A Medieval Czech Penitential Prayer Behind the Cryptographic Enigma of Santa Maria La Nova?
(Tartu University Library, 2026-06-22) Palma, Cosimo; Helm, Louie; Desenclos, Camille; Pierrot, Cécile
This paper presents renewed attack on the encrypted inscription housed in the Turbolo Chapel of the Neapolitan Church of Santa Maria La Nova. Following the cryptanalytic hypothesis of monoalphabetic substitution and the multilingual leads explored in related works, the present contribution investigates Old Czech, a candidate not yet systematically tested against the cipher. Through a combination of n-gram-driven automated decryption via AZdecrypt and LLM-assisted segmentation of the resulting scriptio continua, a candidate reading has been obtained, pointing to a medieval Czech penitential prayer. Although a definitive solution cannot be provided without the support of philologists and cryptologists versed in Czech and related languages, the convergence of statistical, lexicographic and thematic evidence constitutes to date the strongest cue for this long-standing historical puzzle.
listelement.badge.access-status Avatud juurdepääs ,
New records for Playfair solutions
(Tartu University Library, 2025) Bean, Richard; Helm, Louie; Antal, Eugen; Marák, Pavol
We give solutions to the 24 letter and 22 letter Playfair challenges proposed in Dunin et al. (2022) A number of methods were tried combining successful approaches of previous solvers, introducing new ideas while using letter-level and word-level approaches. We used vanilla and positional n-gram models for n values from 6 up to 10. However, these did not greatly assist to distinguish the intended solution from other high-scoring solutions. The most effective discriminative approach involved using a multi-terabyte-scale, unpruned large language model from Buck, Heafield, and Van Ooyen (2014) which moved the solution in each case into the top 5,000 ranked possibilities.
listelement.badge.access-status Avatud juurdepääs ,
Solving a 750-Letter General Bigram Substitution Challenge
(Tartu University Library, 2025) Schmeh, Klaus; Dunin, Elonka; Van Eycke, Jarl; Helm, Louie; Antal, Eugen; Marák, Pavol
The general bigram substitution cipher is an encryption method originating in the Renaissance. It operates using a substitution table that maps each possible letter pair (bigram) to a unique replacement. While conceptually straightforward, this cipher is notably challenging to break, particularly when dealing with short ciphertexts. To inspire further research, one of the authors initiated a bigram substitution challenge featuring a 750-character ciphertext. In this paper, we present the solution to that challenge, achieved by two other authors using a hill climbing algorithm combined with a scoring function based on 8-gram (eight-letter sequence) frequencies. Since no prior 8-gram frequency statistics existed for the English language, one of the authors developed a comprehensive dataset by analyzing 2 terabytes of text, including 5.8 million books and the entire content of Wikipedia. This achievement, to our knowledge, marks the shortest bigram substitution ciphertext ever successfully decrypted. Furthermore, we propose a new challenge based on a 600-character ciphertext and invite readers to tackle it, setting the stage for future advancements in this field.