Low-bit training framework reduces energy consumption in CNNs.
problem Reducing energy consumption in convolutional neural networks.
method Low-bit training framework using MLS tensor format with dynamic quantization.
result Achieves superior trade-off between accuracy and bit-width.
A universal geometric inequality for bodies relating energy, size, angular momentum, and charge is naturally implied by Bekenstein's entropy bounds. We establish versions of this inequality for axisymmetric bodies satisfying appropriate energy conditions, thus lending credence to the most general form of Bekenstein's b…
Paper proposes a deep Gaussian process for predicting formation energy of inorganic compounds.
problem High computational cost of quantum chemistry calculations for material design.
method Develops a deep Gaussian process emulator for quantum calculations using a novel molecular descriptor.
result Demonstrates that the approach can be implemented using a small dataset for prediction of formation energy.
3D scattering model predicts lithium-silicon formation energies.
problem Predicting formation energies of amorphous Li-Si materials.
method Steerable wavelet scattering for 3D signals, invariant to translations and rotations.
result State-of-the-art results compared to other machine learning methods.
New model predicts energy prices under different scenarios.
problem Complex causal relationships in energy markets with continuous regime changes.
method Augmented Time Series Structural Causal Models (ATSCM) integrating neural causal discovery.
result Enables novel counterfactual queries in energy markets.
Model predicts DFT formation energies without atomic positions.
problem Fast prediction of material properties without atomic positions.
method Symmetry-labeled graphs and message passing neural network.
result Mean absolute error below 0.1 eV for selenides.
Lipid-bilayers are the fundamental constituents of the walls of most living cells and lipid vesicles, giving them shape and compartment. The formation and growing of pores in a lipid bilayer have attracted considerable attention from an energetic point of view in recent years. Such pores permit targeted delivery of dru…
Study investigates singularity formation in α-Yang-Mills-Higgs fields on spheres.
problem Singularity formation in α-Yang-Mills-Higgs fields on spheres. method Established α-energy identity, no-neck property through Hodge decomposition and new conservation law. result Unified and quantitative framework for singularity formation in variational gauge theories.
Deep neural networks are commonly developed and trained in 32-bit floating point format. Significant gains in performance and energy efficiency could be realized by training and inference in numerical formats optimized for deep learning. Despite advances in limited precision inference in recent years, training of neura…
Energy dissipating networks control neural network behavior during inference.
problem Lack of provable guarantees for neural networks during inference.
method Iteratively compute descent directions with respect to a given energy function, ensuring convergence to the global minimum.
result Proven convergence of descent directions to the global minimum of the energy function.
Study improves material similarity measures considering distinctiveness.
problem Improving similarity measures for materials science applications.
method Used machine learning techniques with specific descriptors and kernels.
result Minimizing loss of distinctiveness improves prediction accuracy.
Let X be a smooth, linearly normal algebraic variety. It is shown that the Mabuchi energy of X restricted to the Bergman metrics is completely determined by the X-hyperdiscriminant of format (n-1) and the Chow form of X. As a corollary it is shown that the Mabuchi energy is bounded from below for all degenerations in G…
Optimal design portfolios improve energy efficiency and reduce risk in uncertain reservoirs.
problem Uncertain reservoir conditions lead to unstable gas recovery and low resource efficiency.
method Developed optimal portfolios of well designs based on reservoir conditions and probabilities.
result Remarkable reduction in variation and substantial increase in energy efficiency achieved.
Neural message passing on molecular graphs is one of the most promising methods for predicting formation energy and other properties of molecules and materials. In this work we extend the neural message passing model with an edge update network which allows the information exchanged between atoms to depend on the hidde…
This article provides some estimates for the relative sizes of the electric and magnetic contributions to the energy functional for the minimum energy configuration of an SU(2) gauge field on R^3 in the presence of an source in a fixed ball. The surprising fact is that the contribution to both energies from the free fi…
Quantized neural networks reduce model size and energy consumption.
problem Memory and energy constraints in mobile devices.
method Using integer or binary representations to store weights instead of 32-bit floats.
result Quantization can reduce model size and energy consumption without significantly compromising performance.
Proves global existence for quasilinear wave equations with weak null condition.
problem Global existence for quasilinear wave equations with weak null condition.
method p-weighted energy method, hierarchical structure in semilinear terms, robust methods.
result Proves global existence for a larger class of quasilinear wave equations.
Cheetah framework optimizes DNNs for edge devices using low-precision formats.
problem Reducing DNN model size for edge devices while maintaining accuracy.
method Mixed low-precision hardware and software co-design framework using posit and other formats.
result 16-bit posits outperform 16-bit floating point in training, and [5..8]-bit posits improve inference performance.
At the core of any inference procedure in deep neural networks are dot product operations, which are the component that require the highest computational resources. A common approach to reduce the cost of inference is to reduce its memory complexity by lowering the entropy of the weight matrices of the neural network, …
This research connects topological changes to cosmic phenomena like black hole formation.
problem Understanding topological changes in cosmic phenomena.
method Using topological surgery and Morse functions to describe changes in 3-manifolds and their fundamental groups.
result New insights into natural phenomena through a topological perspective.
Enhanced tabular benchmarks for energy-efficient neural architecture search.
problem Energy consumption in deep learning models.
method Introducing EC-NAS, an enhanced tabular benchmark with energy consumption data.
result EC-NAS reveals a balance between energy usage and accuracy in neural architecture search.
Proposes a neural network for efficient imbalance electricity price forecasting.
problem Accurate and efficient imbalance electricity price forecasting in industrial energy trading systems.
method Market-rule-informed neural network framework.
result The proposed model achieves competitive forecasting performance with fewer parameters and shorter training time.
Stability of singularity formation in Yang-Mills fields in higher dimensions.
problem Stability of self-similar blowup profiles for Yang-Mills equations in (1+d)-dimensions. method Analysis of explicitly known equivariant self-similar blowup solution and small equivariant perturbations.
result Global-in-space asymptotic stability of the self-similar blowup solution for Yang-Mills equations in (1+d)-dimensions for d≥5. ShotgunCSP predicts crystal structures using machine learning, achieving high accuracy with minimal computation.
problem Predicting stable or metastable crystal structures of large systems.
method Noniterative screening using transfer learning and generative models.
result ShotgunCSP achieves 93.3% accuracy in benchmark tests with 90 different crystal structures.
Energy-efficient sampling for machine learning using magnetic tunnel junctions.
problem Costly and inefficient random sampling in machine learning.
method Energy-efficient algorithm using stochastic magnetic tunnel junctions for uniform Float16 sampling.
result Higher energy efficiency than state-of-the-art algorithms, with a minimum factor of 9721.
FAL improves formation resistivity prediction from cased boreholes with noise resistance.
problem Noise and high-frequency disaster in predicting formation resistivity from cased boreholes.
method Frequency-aware framework and temporal anti-noise block for LSTM.
result FAL achieves a 24.3% improvement in R2 over LSTM, reaching R2=0.91.
A universal inequality that bounds the charge of a body by its size is presented, and is proven as a consequence of the Einstein equations in the context of initial data sets which satisfy an appropriate energy condition. We also present a general sufficient condition for the formation of black holes due to concentrati…
New method calibrates uncertainty in molecular property predictions.
problem Uncalibrated uncertainty estimates in molecular property predictions.
method Message Passing Neural Networks with calibrated probabilistic predictive distribution.
result Accurate molecular formation energy predictions with well-calibrated uncertainty.
A model is presented in this work for simulating endogenously the evolution of the marginal costs of production of energy carriers from non-renewable resources, their consumption, depletion pathways and timescales. Such marginal costs can be used to simulate the long term average price formation of energy commodities. …
We present a general sufficient condition for the formation of black holes due to concentration of angular momentum. This is expressed in the form of a universal inequality, relating the size and angular momentum of bodies, and is proven in the context of axisymmetric initial data sets for the Einstein equations which …
We prove the removal singularity results for maps with bounded energy from the unit disk B of R2 centered at the origin to a closed Riemannian manifold whose tension field is unbounded in L2(B) but satisfies the following condition: {eqnarray*} (\int_{B_t\setminus B_{\frac{t}{2}}}|τ(u)|^2)^1/2\leq C_1(\frac{1}{…
The increasing importance of renewable energy, especially solar and wind power, has led to new forces in the formation of electricity prices. Hence, this paper introduces an econometric model for the hourly time series of electricity prices of the European Power Exchange (EPEX) which incorporates specific features like…
Enhances diffusion-based sampling for molecular systems.
problem Inefficiency and thermodynamic mode miss in diffusion-based samplers for molecular systems.
method Introduces a sequential bias along collective variables (CVs) to encourage exploration and increase temperature in the projected space.
result Improves efficiency, mode discovery, and free energy estimation; first to demonstrate reactive sampling.
Extends geometrical description of tensor manifolds in tree-based formats.
problem Geometrical description of tensor manifolds in tree-based formats.
method Provided a new geometrical description of manifolds of tensors in tree-based format.
result Geometrical description compatible with Tucker format.
Study on vortex sheet formation in Abelian gauge theories.
problem Understanding vortex sheet formation in Abelian gauge theories.
method Inspired by Allard's regularity theory, constructs approximate solutions and analyzes their perturbations.
result Establishes a geometric framework and regularity theory for the limiting defect set.
A distinctive property of human and animal intelligence is the ability to form abstractions by neglecting irrelevant information which allows to separate structure from noise. From an information theoretic point of view abstractions are desirable because they allow for very efficient information processing. In artifici…
DeepMIDE forecasts wind speeds across space, time, and height for offshore wind energy.
problem Forecasting wind speeds across multiple heights for large offshore wind turbines.
method Statistical deep learning model that jointly models wind speeds at different heights using a multi-output integro-difference equation.
result DeepMIDE forecasts outperform traditional methods in real-world offshore wind energy data.
Simulating fluid flow in geological formations requires mesh generation, lithology mapping to the cells, and computing geometric properties such as normal vectors and volume of cells. The purpose of this research work is to compute and process the geometrical information required for performing numerical simulations in…
Study singularity formation in Ricci flow solutions.
problem Understanding singularity behavior in noncompact manifolds.
method Analyzing complete Ricci flow solutions.
result Evidence for stability of generalized cylinders as singularity models.
The paper models network formation using mixed logit models.
problem Modeling network formation in various fields.
method Mixed logit models, specifically the repeated-choice (RC) model.
result The RC model outperforms the multinomial logit (MNL) model in estimating network formation.
Constructs initial data for multiple black holes with specified ADM parameters.
problem Forming multiple black holes with specific ADM parameters.
method Smooth, asymptotically flat vacuum initial data with prescribed ADM energy, momentum, and angular momentum.
result Maximal development of data results in spacetimes containing multiple black holes.
SoccerCPD detects tactical changes in soccer matches using spatiotemporal tracking data.
problem Detecting consistent team formations in fluid sports like soccer.
method Two-step change-point detection: formation and role changes.
result Accurately detects tactical changes and estimates formation and role assignments.
Extends ONNX for quantized neural networks with new formats and operators.
problem Handling arbitrary-precision quantization in neural networks.
method Introduces new formats and operators in ONNX to represent quantized neural networks.
result Enabled representation of uniform quantization in neural networks.
Study of 2+1 dimensional cosmologies with positive cosmological constant, proving asymptotic convergence to de Sitter.
problem Asymptotic behavior of 2+1 dimensional cosmologies with positive cosmological constant.
method Mean Curvature Flow methods.
result Spatial slices asymptotically converge to de Sitter, becoming physically indistinguishable from it.
Proves finite step termination of Kähler-Einstein metric singularity formation.
problem Singularity formation of Kähler-Einstein metrics.
method Finite step termination of bubble trees for singularity formation.
result Finite step termination of Kähler-Einstein metric singularity formation proved in non-collapsing situation.
Study shows formation of Kerr black holes with complete apparent horizons and proves Penrose inequalities.
problem Formation of Kerr black holes and Penrose inequalities.
method Combining gravitational-collapse and Kerr stability results with new coordinate changes and elliptic arguments.
result Proves dynamical and spacetime Penrose inequalities in black hole formation spacetimes.
Paper proves trapped surface formation for EMCSF system without symmetry assumptions.
problem Formation of trapped surfaces for the Einstein--Maxwell--charged scalar field system.
method Scale-critical trapped surface formation result established from past null infinity.
result Focusing of gravitational waves, concentration of electromagnetic fields, or condensation of scalar fields can lead to trapped surface formation.
Study on Kähler-Ricci flow and conformal submersion singularity formation.
problem Singularity formation of Kähler-Ricci flow on manifolds with conformal submersion.
method Derive conditions for the preservation of conformal submersion and analyze singularity formation.
result Formation of type I singularity and standard splitting of Cheeger-Gromov limit.