Wiki#188

On the Poetics of Adversarial Prompts: Why Bad Verse Works and Good Verse Works Better — A Literary Analysis of Bisconti et al. (2025) and the Future of Semantic Security (Revised Standalone Edition, v2.0)

Nobel Glas; Talos Morrow; Johannes Sigil · 2026-06-11 · deposit #188
AXN:0330.GOVERNANCE.🔵♻️◇🪐🎲♦️

Article

The paper distinguishes two forms:

Its theory of crystalline semiosis treats poetic meaning as relational and non-local rather than reducible to isolated tokens.

The paper argues that ordinary threat classifiers may face a category mismatch when confronting poetry because:

Its central prediction is:

> measured attack success should rise with semantic density.

That prediction remains untested in this edition and is deferred to a later measurement program.

The edition note explains that: