English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent work has shown that despite their impressive capabilities,text-to-image diffusion models such as DALL-E 2 (Ramesh et al., 2022) candisplay strange behaviours when a prompt contains a word with multiple possiblemeanings, often generating images containing both senses of the word (Rassin etal., 2022). In this work we seek to put forward a possible explanation of thisphenomenon. Using the similar Stable Diffusion model (Rombach et al., 2022), wefirst show that when given an input that is the sum of encodings of twodistinct words, the model can produce an image containing both conceptsrepresented in the sum. We then demonstrate that the CLIP encoder used toencode prompts (Radford et al., 2021) encodes polysemous words as asuperposition of meanings, and that using linear algebraic techniques we canedit these representations to influence the senses represented in the generatedimages. Combining these two findings, we suggest that the homonym duplicationphenomenon described by Rassin et al. (2022) is caused by diffusion modelsproducing images representing both of the meanings that are present insuperposition in the encoding of a polysemous word.

Schr\"{o}dinger's Bat: Diffusion Models Sometimes Generate Polysemous Words in Superposition