The paper introduces SymbolicAI, a framework that uses a logic-based approach for concept learning and flow management in generative processes. It integrates generative models with various solvers, treating large language…
Continue readingThis study investigates adversarial attacks through the lens of diffusion models, but not for enhancing the adversarial robustness of image classifiers. The focus is on utilizing the diffusion model to…
Continue readingThe paper discusses the best-of-n policy used for aligning generative models. It disproves a common claim that the KL divergence between the best-of-n policy and the base policy is equal…
Continue readingThis academic article focuses on the evaluation of generative models of expressive piano performance. These models are usually assessed by comparing their predictions to a human performance. The study presents…
Continue readingThe article discusses the use of normalizing flows in voice conversion (VC) tasks and introduces a new training paradigm called AutoEncoder Normalizing Flow (AE-Flow). Normalizing flows are unsupervised generative models…
Continue readingThe study discusses the potential harms posed by AI-generated texts, especially from factoid, unfair, and toxic content. The authors propose FFT, a new benchmark with 2116 instances, for evaluating the…
Continue readingThe paper by Joseph Goodier and Neill D. F. Campbell focuses on Out-of-Distribution (OOD) detection with Denoising Diffusion Probabilistic Models (DDPMs). OOD detection is crucial in machine learning to identify…
Continue readingThe paper presents a novel approach to generative models, extending the continuous time framework from Brownian motion to fractional Brownian motion (FBM). The authors propose Generative Fractional Diffusion Models (GFDM),…
Continue readingThis academic article presents a compressed sensing (CS) setup for the reconstruction of Generative Sparse-Latent (GSL) signals, which are generated by a neural network excited by a sparse-latent input signal….
Continue readingThe article introduces Marginalization Models (MAMs), a new type of generative model for high-dimensional discrete data. MAMs offer scalable and flexible generative modeling with tractable likelihoods. They facilitate fast evaluation…
Continue readingLooks like you're using an ad blocker. Please support us by disabling these ad blocker.
