Zobrazeno 1 - 10
of 31
pro vyhledávání: '"Cangea, Cătălina"'
Reinforcement learning has seen increasing applications in real-world contexts over the past few years. However, physical environments are often imperfect and policies that perform well in simulation might not achieve the same performance when applie
Externí odkaz:
http://arxiv.org/abs/2211.06929
Autor:
Kossen, Jannik, Cangea, Cătălina, Vértes, Eszter, Jaegle, Andrew, Patraucean, Viorica, Ktena, Ira, Tomasev, Nenad, Belgrave, Danielle
We introduce a challenging decision-making task that we call active acquisition for multimodal temporal data (A2MT). In many real-world scenarios, input features are not readily available at test time and must instead be acquired at significant cost.
Externí odkaz:
http://arxiv.org/abs/2211.05039
Autor:
Hawthorne, Curtis, Jaegle, Andrew, Cangea, Cătălina, Borgeaud, Sebastian, Nash, Charlie, Malinowski, Mateusz, Dieleman, Sander, Vinyals, Oriol, Botvinick, Matthew, Simon, Ian, Sheahan, Hannah, Zeghidour, Neil, Alayrac, Jean-Baptiste, Carreira, João, Engel, Jesse
Real-world data is high-dimensional: a book, image, or musical performance can easily contain hundreds of thousands of elements even after compression. However, the most commonly used autoregressive models, Transformers, are prohibitively expensive t
Externí odkaz:
http://arxiv.org/abs/2202.07765
Structure-based drug design involves finding ligand molecules that exhibit structural and chemical complementarity to protein pockets. Deep generative methods have shown promise in proposing novel molecules from scratch (de-novo design), avoiding exh
Externí odkaz:
http://arxiv.org/abs/2111.04107
Neural Processes (NPs) are powerful and flexible models able to incorporate uncertainty when representing stochastic processes, while maintaining a linear time complexity. However, NPs produce a latent description by aggregating independent represent
Externí odkaz:
http://arxiv.org/abs/2009.13895
Autor:
Knyazev, Boris, de Vries, Harm, Cangea, Cătălina, Taylor, Graham W., Courville, Aaron, Belilovsky, Eugene
Inferring objects and their relationships from an image in the form of a scene graph is useful in many applications at the intersection of vision and language. We consider a challenging problem of compositional generalization that emerges in this tas
Externí odkaz:
http://arxiv.org/abs/2007.05756
Autor:
Mernyei, Péter, Cangea, Cătălina
We present Wiki-CS, a novel dataset derived from Wikipedia for benchmarking Graph Neural Networks. The dataset consists of nodes corresponding to Computer Science articles, with edges based on hyperlinks and 10 classes representing different branches
Externí odkaz:
http://arxiv.org/abs/2007.02901
Autor:
Taylor-King, Jake P., Regep, Cristian, Soman, Jyothish, Tong, Flawnson, Cangea, Catalina, Roberts, Charlie
Dynamic Distribution Decomposition (DDD) was introduced in Taylor-King et. al. (PLOS Comp Biol, 2020) as a variation on Dynamic Mode Decomposition. In brief, by using basis functions over a continuous state space, DDD allows for the fitting of contin
Externí odkaz:
http://arxiv.org/abs/2006.05138
Autor:
Knyazev, Boris, de Vries, Harm, Cangea, Cătălina, Taylor, Graham W., Courville, Aaron, Belilovsky, Eugene
Scene graph generation (SGG) aims to predict graph-structured descriptions of input images, in the form of objects and relationships between them. This task is becoming increasingly useful for progress at the interface of vision and language. Here, i
Externí odkaz:
http://arxiv.org/abs/2005.08230
Recent advancements in graph representation learning have led to the emergence of condensed encodings that capture the main properties of a graph. However, even though these abstract representations are powerful for downstream tasks, they are not equ
Externí odkaz:
http://arxiv.org/abs/2002.03864