Related papers: LINNA: Likelihood Inference Neural Network Acceler…

Large reasoning models achieve strong performance on complex tasks by generating extended chains of thought, but they often "overthink": continuing to reason long after they have enough information to answer correctly. This wastes…

In previous works, we proposed to estimate cosmological parameters with the artificial neural network (ANN) and the mixture density network (MDN). In this work, we propose an improved method called the mixture neural network (MNN) to…

We present a comparison of simulation-based inference to full, field-based analytical inference in cosmological data analysis. To do so, we explore parameter inference for two cases where the information content is calculable analytically:…

We study the problem of optimizing Large Language Model (LLM) inference scheduling to minimize total latency. LLM inference is an online and multi-task service process and also heavily energy consuming by which a pre-trained LLM processes…

Recurrent neural networks such as Long Short-Term Memories (LSTMs) learn temporal dependencies by keeping an internal state, making them ideal for time-series problems such as speech recognition. However, the output-to-input feedback…

The inference of cosmological quantities requires accurate and large hydrodynamical cosmological simulations. Unfortunately, their computational time can take millions of CPU hours for a modest coverage in cosmological scales ($\approx (100…

Large language models (LLMs) offer remarkable capabilities, yet their high inference costs restrict wider adoption. While increasing parameter counts improves accuracy, it also broadens the gap between state-of-the-art capabilities and…

Inferring the astrophysical parameters of coalescing compact binaries is a key science goal of the upcoming advanced LIGO-Virgo gravitational-wave detector network and, more generally, gravitational-wave astronomy. However, current…

Long short-term memory (LSTM) is a robust recurrent neural network architecture for learning spatiotemporal sequential data. However, it requires significant computational power for learning and implementing from both software and hardware…

We present a framework that combines physics-informed neural networks (PINNs) with Markov Chain Monte Carlo (MCMC) inference to constrain dynamical dark energy models using the Pantheon+ Type Ia supernova compilation. First, we train a…

The capabilities and adoption of deep neural networks (DNNs) grow at an exhilarating pace: Vision models accurately classify human actions in videos and identify cancerous tissue in medical scans as precisely than human experts; large…

Current ground-based cosmological surveys, such as the Dark Energy Survey (DES), are predicted to discover thousands of galaxy-scale strong lenses, while future surveys, such as the Vera Rubin Observatory Legacy Survey of Space and Time…

Satisfying the high computation demand of modern deep learning architectures is challenging for achieving low inference latency. The current approaches in decreasing latency only increase parallelism within a layer. This is because…

Solving inverse problems with Physics-Informed Neural Networks (PINNs) is computationally expensive for multi-query scenarios, as each new set of observed data requires a new, expensive training procedure. We present Inverse-Parameter Basis…

The sound speed of dark energy encodes fundamental information about the microphysics underlying cosmic acceleration, yet remains essentially unconstrained by existing observations. We demonstrate that a lunar-based laser interferometer,…

Strong lensing gravitational time delays are a powerful and cost effective probe of dark energy. Recent studies have shown that a single lens can provide a distance measurement with 6-7 % accuracy (including random and systematic…

The Laser Interferometer Space Antenna (LISA) will be capable of detecting gravitational waves (GWs) in the milli-Hertz band. Among various sources, LISA will detect the coalescence of supermassive black hole binaries (SMBHBs). Accurate and…

This paper presents novel reconfigurable architectures for reducing the latency of recurrent neural networks (RNNs) that are used for detecting gravitational waves. Gravitational interferometers such as the LIGO detectors capture cosmic…

The emergence of neural network capabilities invariably leads to a significant surge in computational demands due to expanding model sizes and increased computational complexity. To reduce model size and lower inference costs, recent…

Integrated Nested Laplace Approximations (INLA) has been a successful approximate Bayesian inference framework since its proposal by Rue et al. (2009). The increased computational efficiency and accuracy when compared with sampling-based…