Multi Layer

Combined approaches that stack multiple encryption steps. For example, replacing letters first, then scrambling their order. Includes filler-letter removal, cascaded layers, and joint optimization.

What we tested: Combinations of two or more layers are one of the main remaining search areas for K4. We have tested a large but bounded part of that space, including many structured rearrangement methods paired with keyed substitution, and found no signal. Much remains open: combinations beyond the grid widths, key lengths and alphabets searched so far; self-keying (autokey) ciphers combined with a letter-rearrangement step, which the known letters alone cannot rule out; and non-standard procedures or key sources.

73 eliminations recorded.

Title	Configs	Best	Verdict	Tier
Vigenère-family substitution before or after columnar transposition (two-layer test) Tested a Vigenère-family substitution followed by a columnar transposition, and the reverse order, using keys worked out from the known letters and a …	109,899,094	9/24	NOISE	N/A
Hill n=2, then keyed columnar w2-10 Tests invertible Hill cipher n=2 followed by keyed columnar transposition, widths 2-10, with the known plaintext at its carved positions. Every settin…	64,606,592	N/A	ELIMINATED	Tier 2
Running key from the K1-K3 plaintexts + width-8 columnar transposition A running key is a long passage of text used as the key. A complete re-run on 2026-09-18 tried 15 passages built from the K1, K2 and K3 plaintexts, at…	52,000,000	7/24	NOISE	Tier 2
Column and route rearrangements plus a repeating key, on 73 letters left by a column-band filler model A repeating-key Vigenere, Beaufort or Variant Beaufort cipher (key lengths 1 to 13) combined with column, rail-fence and route rearrangements, on the …	31,800,000	18/24	NOISE	Tier 2
Transposition-guided running key fragment scoring (28M Monte Carlo, 28 cores) Transposition-guided running key fragment scoring — optimal assignment via Monte Carlo (28M samples, 28 cores), using a combined approach: scramble th…	28,000,000	N/A	NOISE	Tier 3
Ciphertext feedback, then a frozen route bank Tests signed ciphertext-feedback keys with lags 1-25 followed by the frozen transposition route bank, with the known plaintext at its carved positions…	21,289,600	N/A	ELIMINATED	Tier 2
Repeating-key cipher followed by a columnar transposition, widths 4 to 9, every column order Tried a repeating-key cipher followed by a columnar transposition, at widths 4 to 9 in every column order and key lengths 1 to 26. No setting with a k…	21,273,408	N/A	NOISE	Tier 2
Keyed columnar w2-10 plus any monoalphabetic key Tests keyed columnar transposition, widths 2-10 (every column order, both directions), combined with any bijective monoalphabetic key, with the known …	8,075,824	N/A	ELIMINATED	Tier 2
Unit-affine feedback, then a frozen route bank Tests unit-affine ciphertext-feedback keys with lags 1-31 followed by the frozen transposition route bank, with the known plaintext at its carved posi…	6,599,776	N/A	ELIMINATED	Tier 2
Unit-multiplier quadratic keys, then a route bank Tests unit-multiplier quadratic key progression (periods 1, 2, 3, 4, 6) followed by the frozen transposition route bank, with the known plaintext at i…	6,386,880	N/A	ELIMINATED	Tier 2
Unit-multiplier cubic keys, then a route bank Tests unit-multiplier cubic key progression (periods 1-4) followed by the frozen transposition route bank, with the known plaintext at its carved posi…	5,109,504	N/A	ELIMINATED	Tier 2
8-row grid transposition (from a misreading of a Sanborn note) Tested columnar grids of width 12, 13 and 14, filled row by row, with Vigenère-family substitution, using a combined approach: scramble the letter ord…	4,500,000	N/A	NOISE	N/A
Keyed columnar w2-8 plus unit-affine letter map Tests keyed columnar transposition, widths 2-8, combined with a constant unit-affine letter map, with the known plaintext at its carved positions. Eve…	4,438,272	N/A	ELIMINATED	Tier 2
Mitm Mono Transposition Tested a combined approach: scramble the letter order AND replace each letter. 4.2 million key/parameter combinations were tested.	4,209,541	N/A	NOISE	N/A
Mitm Periodic Transposition Tested a combined approach: scramble the letter order AND replace each letter. 4.2 million key/parameter combinations were tested.	4,209,541	N/A	NOISE	N/A
Self-keying (autokey) cipher after structured letter rearrangements (consistency filter) Tried self-keying ciphers after about 4.2 million structured letter rearrangements. For the plaintext-keyed form only a handful of known letters can b…	4,207,963	N/A	NOISE	Tier 3
Blitz Strip Cipher V2 Tested a combined approach: scramble the letter order AND replace each letter. 3.6 million key/parameter combinations were tested.	3,608,555	N/A	NOISE	N/A
Linear progressive key, then a frozen route bank Tests linear progressive key followed by the frozen transposition route bank, with the known plaintext at its carved positions. Every setting in scope…	2,554,752	N/A	ELIMINATED	Tier 2
Targeted search for transpositions resolving homophonic contradictions Tested a method where each letter can be represented by multiple different symbols. 1.3 million key/parameter combinations were tested.	1,282,433	N/A	NOISE	N/A
Quadratic key progression, then a frozen route bank Tests quadratic key progression (periods 1, 2, 3, 4, 6) followed by the frozen transposition route bank, with the known plaintext at its carved positi…	1,064,480	N/A	ELIMINATED	Tier 2
Cubic key progression, then a frozen route bank Tests cubic key progression (periods 1-4) followed by the frozen transposition route bank, with the known plaintext at its carved positions. Every set…	851,584	N/A	ELIMINATED	Tier 2
Keyed columnar w2-9 plus any monoalphabetic key Tests keyed columnar transposition, widths 2-9, combined with any bijective monoalphabetic key, with the known plaintext at its carved positions. Ever…	818,224	N/A	ELIMINATED	Tier 2
Hill or affine Hill, then a frozen route bank Tests Hill or affine Hill cipher followed by each of 26,612 frozen transposition routes, with the known plaintext at its carved positions. Every setti…	532,240	N/A	ELIMINATED	Tier 2
Quartic key progression, then a frozen route bank Tests quartic key progression (periods 1 and 2) followed by the frozen transposition route bank, with the known plaintext at its carved positions. Eve…	425,792	N/A	ELIMINATED	Tier 2
First run of the clue-bounded sweep-authoring solver against real K4: 484 machine-authored bounded specs ov... Er two- and three-layer compositions, direct positional alignment, using a combined approach: scramble the letter order AND replace each letter. 250 t…	250,348	7/24	ELIMINATED	N/A
Free-alignment extension of the clue-bounded solver run: the same 484 machine-authored specs rescored with... Known-plaintext words allowed to land anywhere (free crib alignment), three-layer arms included, using a combined approach: scramble the letter order …	250,348	N/A	ELIMINATED	N/A
Blitz Strip Cipher Tested a combined approach: scramble the letter order AND replace each letter. 118 thousand key/parameter combinations were tested.	118,317	N/A	NOISE	N/A
Monoalphabetic + transposition + running key degrees of freedom analysis Tested a combined approach: scramble the letter order AND replace each letter. 100 thousand key/parameter combinations were tested.	100,000	N/A	NOISE	Tier 3
Keyed columnar w2-8 plus any monoalphabetic key Tests keyed columnar transposition, widths 2-8, combined with any bijective monoalphabetic key, with the known plaintext at its carved positions. Ever…	92,464	N/A	ELIMINATED	Tier 2
Partitioned encryption — two systems on two message parts Tested multiple encryption steps applied one after another. 75 thousand key/parameter combinations were tested. The best result matched all 24 known l…	75,168	24/24	NOISE	N/A
Four 6x4 grids stacked with independent shifts. Palette overlap = null. DEFECTOR/ABSCISSA AZ_beau + col6/col7. Four 6x4 grids stacked with independent shifts, using the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to d…	27,848	N/A	NOISE	N/A
Inclinare Stacked Null Mask Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message. 4 thousand key/parame…	4,313	6/24	ELIMINATED	N/A
Null cipher / skip cipher / steganographic extraction patterns On K4 CT, using the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message. Every combin…	2,500	N/A	NOISE	Tier 2
Novel multi-stage decryption pipelines Tested multiple encryption steps applied one after another. 2 thousand key/parameter combinations were tested.	2,462	N/A	NOISE	N/A
Grid reading orders plus a repeating key, on 73 letters left by a column-band filler model More than ten ways of reading a 28-by-31 layout of the cipher text, each with a repeating key of length 1 to 26, on the 73 letters left by a column-ba…	2,262	22/24	NOISE	Tier 2
Palette Exhaustive Null Mask Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message. 8 key/parameter combi…	8	N/A	NOISE	N/A
Blitz CSP Structural Tested backward reasoning from known constraints to narrow possibilities.	N/A	N/A	NOISE	N/A
CSP P23 W15 Beau Tested backward reasoning from known constraints to narrow possibilities. The best result matched all 24 known letters, but on checking it is not a so…	N/A	24/24	NOISE	N/A
S 20 Constraint Propagation Tested backward reasoning from known constraints to narrow possibilities.	N/A	N/A	NOISE	N/A
Egyptological running-key texts with a columnar transposition (widths 5 to 11) Egypt corpus x columnar transposition sweep (widths 5-11, 3 variants, 9 corpora), using a combined approach: scramble the letter order AND replace eac…	N/A	N/A	NOISE	Tier 3
Columnar widths 4, 8, 9 and 10 with a substitution step (from a misread archive note) Tested columnar transposition at widths 4, 8, 9 and 10 combined with a letter-substitution step, using a combined approach: scramble the letter order …	N/A	N/A	NOISE	N/A
Two Layer Stego Cipher Tested a combined approach: scramble the letter order AND replace each letter.	N/A	N/A	NOISE	N/A
K2's cipher (Vigenère-style on the KRYPTOS alphabet, key ABSCISSA) after two width-10 column orders Applied K2's own cipher settings after two width-10 columnar orders kept by an earlier run (f_archive_col_notation_v1), using a combined approach: scr…	N/A	N/A	NOISE	N/A
Brute-force 7 remaining null positions (consensus 17 fixed) Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Bruteforce 7remaining Task3 Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Bruteforce 7remaining Tasks12 Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message. Best match: 15/24 kno…	N/A	15/24	NOISE	N/A
Col8 Null Mask Test Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Extended brute force pos 34-41 with AP + restricted alphabet filters Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
K4 in 4x24 grid (matching K3 chart rows), wheel rotation separates stego Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Morse Binary Null Mask Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Null cipher extraction (E-S-35) Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Stego Cipher Brute Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Stego Oracle Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
All 2,629,575 ways to pick 24 filler letters from the retired filler-palette positions Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
NDYAHR Corrected Neighbor Deletion + LAYER TWO Investigation Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
NDYAHR Neighbor Deletion Layer Two Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
NDYAHR letters in 28x31 grid are directional pointers to neighbors for deletion Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	2/24	NOISE	N/A
Phase 1: Local constant-difference constraint census at all 17 null positions Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Null Mask L0 Breakdown Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Null Mask Ceiling Sweep Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Null Mask CSP Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Null Mask Diversity Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Null Mask Provenance Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Control tests for the retired filler-palette idea Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Stego Joint Mask Solver Universe Audit Tested the hypothesis that some letters in the carved text are meaningless filler (nulls) inserted to disguise the real message.	N/A	N/A	NOISE	N/A
Check that proposed filler-letter masks keep the known letters (not a decryption test) Checks whether the 24 known letters survive when different sets of carved letters are treated as filler and removed, using the hypothesis that some le…	N/A	N/A	NOISE	N/A
Partitioned two-system encryption (variant 2) Tested multiple encryption steps applied one after another. A consistency check this search used to discard candidates was later found invalid for it,…	N/A	N/A	NOISE	N/A
Team Layered Verify Tested multiple encryption steps applied one after another.	N/A	N/A	NOISE	N/A
Team Mono Trans SA Tested multiple encryption steps applied one after another. The best result matched all 24 known letters, but on checking it is not a solution (a matc…	N/A	24/24	NOISE	N/A
K4 frequency distribution analysis and homophonic solver Tested a method where each letter can be represented by multiple different symbols.	N/A	N/A	NOISE	N/A
Systematic search for transpositions resolving all 9 homophonic contradictions Tested a method where each letter can be represented by multiple different symbols.	N/A	N/A	NOISE	N/A
Homophonic + transposition contradiction resolution Tested a method where each letter can be represented by multiple different symbols.	N/A	N/A	NOISE	N/A
Nomenclator + superencipherment model analysis Tested a method where each letter can be represented by multiple different symbols.	N/A	N/A	NOISE	N/A