Skip to content
VALEXII.AI
VALEXII Research / Manuscript tracks

Ancient scripts. Open questions.

A plausible reading can be captivating—and wrong. We begin with reliable records, testable patterns and rival explanations. No track below has a verified semantic decipherment.

RESEARCH GRAPHIC · Original schematic, not manuscript imagery or a translation
123

01 / Corpus control

Preserve transcription and source identity

02 / Structure

Test segmentation, local order and matched nulls

03 / Reconstruction

Held-out prediction, reverse encoding, replication

Semantic status: HOLD / NOT ESTABLISHED · Evidence attachment required

Current tracks

Milestones, not a completion percentage.

Conceptual botanical manuscript with invented markings, inspired by Voynich research.
Voynich · artistic illustration, not an artifact reproduction or proposed reading.
ACTIVE / HOLD

Voynich

Compare transcriptions, boundaries and local patterns against matched random baselines. No semantic decipherment.

Explore
Conceptual Minoan clay tablet with invented marks, inspired by Linear A research.
Linear A · artistic illustration, not an artifact reproduction or proposed reading.
RESEARCH / NOT ESTABLISHED

Linear A

Build reliable corpora and test structural hypotheses. Specific findings need attached evidence before public promotion.

Explore
Conceptual carved wooden glyph board with invented marks, inspired by Rongorongo research.
Rongorongo · artistic illustration, not an artifact reproduction or proposed reading.
RESEARCH / NOT ESTABLISHED

Rongorongo

Study object context and 3D records, test Mamari lunar-calendar alignment, and check fixed values on held-out material. No verified translation.

Explore

Held-out semantic prediction, continuous decoding, reverse reconstruction, and independent replication remain required before any translation claim.