arXiv:2412.06771v3 Announce Type: replace-cross 
Abstract: User prompts for generative AI models are often underspecified, leading to a misalignment between the user intent and models' understanding. As a result, users commonly have to painstakingly refine their prompts. We study this alignment problem in text-to-image (T2I) generation and propose a prototype for proactive T2I agents equipped with an interface to (1) actively ask clarification questions when uncertain, and (2) present their uncertainty about user intent as an understandable and editable belief graph. We build simple prototypes for such agents and propose a new scalable and automated evaluation approach using two agents, one with a ground truth intent (an image) while the other tries to ask as few questions as possible to align with the ground truth. We experiment over three image-text datasets: ImageInWords (Garg et al., 2024), COCO (Lin et al., 2014) and DesignBench, a benchmark we curated with strong artistic and design elements. Experiments over the three datasets demonstrate the proposed T2I agents' ability to ask informative questions and elicit crucial information to achieve successful alignment with at least 2 times higher VQAScore (Lin et al., 2024) than the standard T2I generation. Moreover, we conducted human studies and observed that at least 90% of human subjects found these agents and their belief graphs helpful for their T2I workflow, highlighting the effectiveness of our approach. Code and DesignBench can be found at https://github.com/google-deepmind/proactive_t2i_agents.

تسلط دراسة جديدة الضوء على التحديات التي يواجهها المستخدمون مع نماذج الذكاء الاصطناعي التوليدية، خاصة في توليد الصور من النصوص حيث يمكن أن تكون المطالبات غير واضحة. لمعالجة ذلك، يقترح الباحثون نموذجًا لوكلاء استباقيين يمكنهم طرح أسئلة توضيحية عند عدم اليقين، مما يعزز تجربة المستخدم ويحسن التوافق بين نية المستخدم وفهم النموذج. هذه الابتكار مهم لأنه يمكن أن يبسط العملية الإبداعية للمستخدمين، مما يجعل أدوات الذكاء الاصطناعي أكثر سهولة وفعالية.

Un nuevo estudio destaca los desafíos que enfrentan los usuarios con los modelos de IA generativa, especialmente en la generación de imágenes a partir de texto, donde las indicaciones a menudo pueden ser poco claras. Para abordar esto, los investigadores proponen un prototipo de agentes proactivos que pueden hacer preguntas de aclaración cuando tienen dudas, mejorando la experiencia del usuario y la alineación entre la intención del usuario y la comprensión del modelo. Esta innovación es significativa, ya que podría agilizar el proceso creativo para los usuarios, haciendo que las herramientas de IA sean más intuitivas y efectivas.

Une nouvelle étude met en lumière les défis auxquels les utilisateurs sont confrontés avec les modèles d'IA générative, en particulier dans la génération d'images à partir de texte, où les invites peuvent souvent être floues. Pour y remédier, les chercheurs proposent un prototype d'agents proactifs capables de poser des questions de clarification en cas d'incertitude, améliorant ainsi l'expérience utilisateur et l'alignement entre l'intention de l'utilisateur et la compréhension du modèle. Cette innovation est importante car elle pourrait simplifier le processus créatif pour les utilisateurs, rendant les outils d'IA plus intuitifs et efficaces.

A new study highlights the challenges users face with generative AI models, particularly in text-to-image generation where prompts can often be unclear. To address this, researchers propose a prototype for proactive agents that can ask clarifying questions when uncertain, enhancing user experience and improving alignment between user intent and model understanding. This innovation is significant as it could streamline the creative process for users, making AI tools more intuitive and effective.

Proactive Agents for Multi-Turn Text-to-Image Generation Under Uncertainty

arXiv:2511.18684v1 Announce Type: new 
Abstract: Robust concept removal for text-to-image (T2I) and text-to-video (T2V) models is essential for their safe deployment. Existing methods, however, suffer from costly retraining, inference overhead, or vulnerability to adversarial attacks. Crucially, they rarely model the latent semantic overlap between the target erase concept and surrounding content -- causing collateral damage post-erasure -- and even fewer methods work reliably across both T2I and T2V domains. We introduce Instant Concept Erasure (ICE), a training-free, modality-agnostic, one-shot weight modification approach that achieves precise, persistent unlearning with zero overhead. ICE defines erase and preserve subspaces using anisotropic energy-weighted scaling, then explicitly regularises against their intersection using a unique, closed-form overlap projector. We pose a convex and Lipschitz-bounded Spectral Unlearning Objective, balancing erasure fidelity and intersection preservation, that admits a stable and unique analytical solution. This solution defines a dissociation operator that is translated to the model's text-conditioning layers, making the edit permanent and runtime-free. Across targeted removals of artistic styles, objects, identities, and explicit content, ICE efficiently achieves strong erasure with improved robustness to red-teaming, all while causing only minimal degradation of original generative abilities in both T2I and T2V models.

قدم الباحثون تقنية Instant Concept Erasure (ICE)، وهي نهج جديد لإزالة المفاهيم بشكل موثوق في نماذج النص إلى صورة (T2I) والنص إلى فيديو (T2V). تلغي هذه الطريقة الحاجة إلى إعادة التدريب المكلفة وتقلل من عبء الاستدلال، بينما تعالج الثغرات الأمنية ضد الهجمات العدائية. تستخدم ICE تقنية تعديل الوزن في خطوة واحدة دون تدريب، مما يضمن إلغاء تعلم دقيق ودائم دون أضرار جانبية للمحتوى المحيط.

Investigadores han presentado el Instant Concept Erasure (ICE), un enfoque novedoso para la eliminación robusta de conceptos en modelos de texto-a-imagen (T2I) y texto-a-video (T2V). Este método elimina la necesidad de reentrenamiento costoso y minimiza la sobrecarga de inferencia, al tiempo que aborda las vulnerabilidades a ataques adversariales. ICE emplea una técnica de modificación de pesos en una sola vez, sin necesidad de entrenamiento, que asegura un desaprendizaje preciso y persistente sin daños colaterales al contenido circundante.

Des chercheurs ont introduit l'Instant Concept Erasure (ICE), une nouvelle approche pour la suppression robuste de concepts dans les modèles de texte-à-image (T2I) et de texte-à-vidéo (T2V). Cette méthode élimine le besoin de réentraînement coûteux et minimise la surcharge d'inférence tout en abordant les vulnérabilités aux attaques adversariales. ICE utilise une technique de modification de poids en une seule fois, sans formation, qui garantit un désapprentissage précis et persistant sans dommages collatéraux au contenu environnant.

Researchers have introduced Instant Concept Erasure (ICE), a novel approach for robust concept removal in text-to-image (T2I) and text-to-video (T2V) models. This method eliminates the need for costly retraining and minimizes inference overhead while addressing vulnerabilities to adversarial attacks. ICE employs a training-free, one-shot weight modification technique that ensures precise and persistent unlearning without collateral damage to surrounding content.

Proactive Agents for Multi-Turn Text-to-Image Generation Under Uncertainty

Was this article worth reading? Share it

Augmeta

FLUX AI ART

TypeThinkAI