Search

Vector graphics are essential in design, providing artists with a versatile medium for creating resolution-independent and highly editable visual content. Recent advancements in vision-language and diffusion models have fueled interest in text-to-vector graphics generation. However, existing approaches often suffer from over-parameterized outputs or treat the layered structure - a core feature of vector graphics - as a secondary goal, diminishing their practical use. Recognizing the importance of layered SVG representations, we propose NeuralSVG, an implicit neural representation for generating vector graphics from text prompts. Inspired by Neural Radiance Fields (NeRFs), NeuralSVG encodes the entire scene into the weights of a small MLP network, optimized using Score Distillation Sampling (SDS). To encourage a layered structure in the generated SVG, we introduce a dropout-based regularization technique that strengthens the standalone meaning of each shape. We additionally demonstrate that utilizing a neural representation provides an added benefit of inference-time control, enabling users to dynamically adapt the generated SVG based on user-provided inputs, all with a single learned representation. Through extensive qualitative and quantitative evaluations, we demonstrate that NeuralSVG outperforms existing methods in generating structured and flexible SVG., Comment: Project Page: https://sagipolaczek.github.io/NeuralSVG/

Drug discovery typically consists of multiple steps, including identifying a target protein key to a disease's etiology, validating that interacting with this target could prevent symptoms or cure the disease, discovering a small molecule or biologic therapeutic to interact with it, and optimizing the candidate molecule through a complex landscape of required properties. Drug discovery related tasks often involve prediction and generation while considering multiple entities that potentially interact, which poses a challenge for typical AI models. For this purpose we present MAMMAL - Molecular Aligned Multi-Modal Architecture and Language - a method that we applied to create a versatile multi-task multi-align foundation model that learns from large-scale biological datasets (2 billion samples) across diverse modalities, including proteins, small molecules, and genes. We introduce a prompt syntax that supports a wide range of classification, regression, and generation tasks. It allows combining different modalities and entity types as inputs and/or outputs. Our model handles combinations of tokens and scalars and enables the generation of small molecules and proteins, property prediction, and transcriptomic lab test predictions. We evaluated the model on 11 diverse downstream tasks spanning different steps within a typical drug discovery pipeline, where it reaches new SOTA in 9 tasks and is comparable to SOTA in 2 tasks. This performance is achieved while using a unified architecture serving all tasks, in contrast to the original SOTA performance achieved using tailored architectures. The model code and pretrained weights are publicly available at https://github.com/BiomedSciAI/biomed-multi-alignment and https://huggingface.co/ibm/biomed.omics.bl.sm.ma-ted-458m.

Bioactivity data plays a key role in drug discovery and repurposing. The resource-demanding nature of \textit{in vitro} and \textit{in vivo} experiments, as well as the recent advances in data-driven computational biochemistry research, highlight the importance of \textit{in silico} drug target interaction (DTI) prediction approaches. While numerous large public bioactivity data sources exist, research in the field could benefit from better standardization of existing data resources. At present, different research works that share similar goals are often difficult to compare properly because of different choices of data sources and train/validation/test split strategies. Additionally, many works are based on small data subsets, leading to results and insights of possible limited validity. In this paper we propose a way to standardize and represent efficiently a very large dataset curated from multiple public sources, split the data into train, validation and test sets based on different meaningful strategies, and provide a concrete evaluation protocol to accomplish a benchmark. We analyze the proposed data curation, prove its usefulness and validate the proposed benchmark through experimental studies based on an existing neural network model.

The Internet of Production (IoP) leverages concepts such as digital shadows, data lakes, and a World Wide Lab (WWL) to advance today's production. Consequently, it requires a technical infrastructure that can support the agile deployment of these concepts and corresponding high-level applications, which, e.g., demand the processing of massive data in motion and at rest. As such, key research aspects are the support for low-latency control loops, concepts on scalable data stream processing, deployable information security, and semantically rich and efficient long-term storage. In particular, such an infrastructure cannot continue to be limited to machines and sensors, but additionally needs to encompass networked environments: production cells, edge computing, and location-independent cloud infrastructures. Finally, in light of the envisioned WWL, i.e., the interconnection of production sites, the technical infrastructure must be advanced to support secure and privacy-preserving industrial collaboration. To evolve today's production sites and lay the infrastructural foundation for the IoP, we identify five broad streams of research: (1) adapting data and stream processing to heterogeneous data from distributed sources, (2) ensuring data interoperability between systems and production sites, (3) exchanging and sharing data with different stakeholders, (4) network security approaches addressing the risks of increasing interconnectivity, and (5) security architectures to enable secure and privacy-preserving industrial collaboration. With our research, we evolve the underlying infrastructure from isolated, sparsely networked production sites toward an architecture that supports high-level applications and sophisticated digital shadows while facilitating the transition toward a WWL., Comment: Published in: Brecher, C., Schuh, G., van der Aalst, W., Jarke, M., Piller, F.T., Padberg, M. (eds) Internet of Production. Interdisciplinary Excellence Accelerator Series. Springer, Cham

We propose a new method for obtaining an effective Friedmann-Lema\^itre-Robertson-Walker (FLRW) cosmology from the quantum gravity dynamics of group field theory (GFT), based on the idea that an FLRW universe is characterised by a few macroscopic observables. Rather than relying on assuming a particular type of quantum state and computing expectation values in such a state, here we directly start from relations between macroscopic observables (defined as one-body operators) and formulate dynamics only for those observables. We apply the effective approach to constrained quantum systems (as developed by Bojowald and collaborators) to GFT, providing a systematic expansion in powers of $\hbar$. We obtain a kinematical phase space of expectation values and moments, which does not require an a priori choice of clock variable. We identify a gauge fixing of the system which corresponds to choosing one of the cosmological variables (with the role of extrinsic curvature) as a clock and which allows us to rewrite the effective dynamics in relational form. We show necessary and sufficient conditions for the resulting dynamics of expectation values to be compatible with those of classical FLRW cosmology and discuss the impact of quantum fluctuations., Comment: 33 pages; v2: minor changes in presentation, added references, accepted for publication in Classical and Quantum Gravity

One approach to defining dynamics for quantum gravity in a naturally timeless setting is to select a suitable matter degree of freedom as a 'clock' before quantisation. This idea of deparametrisation was recently introduced in group field theory leading to a Hamiltonian formulation in which states or operators evolve with respect to the clock given by a free massless scalar field, similar to what happens in deparametrised models in quantum cosmology. Here we extend the construction of Hamiltonian group field theory to models with multiple scalar matter fields, encountering new features and technical subtleties compared with the previously studied case. We show that the effective cosmological dynamics for these more general models reduce to the Friedmann dynamics of general relativity with multiple scalar fields in the limit of large volume, if suitable (non-generic) initial conditions are chosen. At high energy, we find corrections to the classical Friedmann equations whose form is similar to what is found in loop quantum cosmology. These corrections lead to generic singularity resolution by a bounce. For generic initial conditions, the effective cosmological dynamics treat the 'clock' field and other matter fields differently, in disagreement with the Friedmann dynamics of general relativity. We speculate on a possible interpretation in terms of inhomogeneities., Comment: 48 pages, 1 figure; v4: Fix typos

We extend various recent results regarding the derivation of effective cosmological Friedmann equations from the dynamics of group field theory (GFT). Restricting ourselves to a single GFT field mode (or fixed values of Peter-Weyl representation labels), we first consider dynamics given by a quadratic Hamiltonian, which takes the form of a squeezing operator, and then add a quartic interaction that can be seen as a toy model for interactions in full GFT. Our derivation of effective Friedmann equations does not require a mean-field approximation; we mostly follow a general approach in which these equations in fact hold for any state. The resulting cosmological equations exhibit corrections to classical Friedmann dynamics similar to those of loop quantum cosmology, leading to generic singularity resolution, but also involve further state-dependent terms. We then specify these equations to various types of coherent states, such as Fock coherent states or Perelomov-Gilmore states based on the su(1,1) structure of harmonic quantum cosmology. We compute relative uncertainties of volume and energy in these states, clarifying whether they can be interpreted as semiclassical. In the interacting case, both analytical and numerical approximations are used to obtain modified cosmological dynamics. Our results clarify how effective cosmological equations derived from GFT can provide reliable approximations to the full dynamics., Comment: 38 pages, 5 figures; v2: improved discussion, added references, additional figure

In this addendum to [Phys. Rev. D 99 (2019) 086017], we extend the construction of a Hamiltonian formalism to a new class of group field theory actions with a kinetic term that is local in the group variables and depends only on their derivatives; such a kinetic term will couple magnetic indices of opposite sign in the Peter-Weyl decomposition. The main results of [Phys. Rev. D 99 (2019) 086017] for the resulting cosmology extend to this case., Comment: 4 pages, addendum to arXiv:1810.01259