Newest papers

Beyond the Prompt: Recovering Hidden Concepts in Text-to-Image Models

Zhengfeng Zheng, Minghong Fang · 2026 · arXiv

DOI: 10.48550/arXiv.2610.27473 · open-access PDF

Spanish

Los modelos de texto a imagen pueden generar conceptos visuales que sus prompts nunca mencionan, y el trabajo reciente muestra que tales asociaciones se pueden implantar de modo que los prompts ordinarios los desencadenen. Auditar este comportamiento requiere recuperar la descripción que el prompt omite. Los optimizadores de prompts existentes no son adecuados para esta tarea. Los métodos basados en gradientes requieren acceso retroactivo al codificador de puntuación. La búsqueda evolutiva discreta trata a los tokens semánticamente no relacionados como igualmente cercanos. La búsqueda guiada por el modelo de idioma requiere una llamada de inferencia por candidato, lo que limita la exploración.

Machine translation (Meta M2M100); the English original follows.

Abstract

Text-to-image models can generate visual concepts that their prompts never mention, and recent work shows that such associations can be implanted so that ordinary prompts trigger them. Auditing this behavior requires recovering the description that a prompt omits. Existing prompt optimizers are not well suited to this task. Gradient-based methods require backward access to the scoring encoder. Discrete evolutionary search treats semantically unrelated tokens as equally close. Language-model-guided search requires one inference call per candidate, which limits exploration. This paper presents E

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.27473, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange