arXiv subjects

Jack O'Brien

Publications and source records attributed to Jack O'Brien.

5 recordsLinked to original sources

arXiv2026

SELDON: Supernova Explosions Learned by Deep ODE Networks

The discovery rate of optical transients will explode to 10 million public alerts per night once the Vera C. Rubin Observatory's Legacy Survey of Space and Time comes online, overwhelming the traditional physics-based inference pipelines. A continuous-time forecasting AI model is of interest because it can deliver millisecond-scale inference for thousands of objects per day, whereas legacy MCMC codes need hours per object. In this paper, we propose SELDON, a new continuous-time variational autoencoder for panels of sparse and irregularly time-sampled (gappy) astrophysical light curves that are nonstationary, heteroscedastic, and inherently dependent. SELDON combines a masked GRU-ODE encoder with a latent neural ODE propagator and an interpretable Gaussian-basis decoder. The encoder learns to summarize panels of imbalanced and correlated data even when only a handful of points are observed. The neural ODE then integrates this hidden state forward in continuous time, extrapolating to future unseen epochs. This extrapolated time series is further encoded by deep sets to a latent distribution that is decoded to a weighted sum of Gaussian basis functions, the parameters of which are physically meaningful. Such parameters (e.g., rise time, decay rate, peak flux) directly drive downstream prioritization of spectroscopic follow-up for astrophysical surveys. Beyond astronomy, the architecture of SELDON offers a generic recipe for interpretable and continuous-time sequence modeling in any time domain where data are multivariate, sparse, heteroscedastic, and irregularly spaced.

astro-ph.IM↗

arXiv2025

Beyond Context Limits: Subconscious Threads for Long-Horizon Reasoning

To break the context limits of large language models (LLMs) that bottleneck reasoning accuracy and efficiency, we propose the Thread Inference Model (TIM), a family of LLMs trained for recursive and decompositional problem solving, and TIMRUN, an inference runtime enabling long-horizon structured reasoning beyond context limits. Together, TIM hosted on TIMRUN supports virtually unlimited working memory and multi-hop tool calls within a single language model inference, overcoming output limits, positional-embedding constraints, and GPU-memory bottlenecks. Performance is achieved by modeling natural language as reasoning trees measured by both length and depth instead of linear sequences. The reasoning trees consist of tasks with thoughts, recursive subtasks, and conclusions based on the concept we proposed in Schroeder et al, 2025. During generation, we maintain a working memory that retains only the key-value states of the most relevant context tokens, selected by a rule-based subtask-pruning mechanism, enabling reuse of positional embeddings and GPU memory pages throughout reasoning. Experimental results show that our system sustains high inference throughput, even when manipulating up to 90% of the KV cache in GPU memory. It also delivers accurate reasoning on mathematical tasks and handles information retrieval challenges that require long-horizon reasoning and multi-hop tool use.

cs.CL↗

arXiv2023

StaNdaRT: A repository of standardized test models and outputs for supernova radiative transfer

We present the first results of a comprehensive supernova (SN) radiative-transfer (RT) code-comparison initiative (StaNdaRT), where the emission from the same set of standardized test models is simulated by currently-used RT codes. A total of ten codes have been run on a set of four benchmark ejecta models of Type Ia supernovae. We consider two sub-Chandrasekhar-mass ($M_\mathrm{tot} = 1.0$ M$_\odot$) toy models with analytic density and composition profiles and two Chandrasekhar-mass delayed-detonation models that are outcomes of hydrodynamical simulations. We adopt spherical symmetry for all four models. The results of the different codes, including the light curves, spectra, and the evolution of several physical properties as a function of radius and time, are provided in electronic form in a standard format via a public repository. We also include the detailed test model profiles and several python scripts for accessing and presenting the input and output files. We also provide the code used to generate the toy models studied here. In this paper, we describe in detail the test models, radiative-transfer codes and output formats and provide access to the repository. We present example results of several key diagnostic features.

astro-ph.SR↗

arXiv2022

Interplay of symmetry-conserved tunneling, interfacial oxidation and perpendicular magnetic anisotropy in CoFeB/MgO-based junctions

The interfacial oxidation level and thermodynamic properties of the MgO-based perpendicular magnetic tunneling junctions are investigated. The symmetry-conserved tunneling effect depends sensitively on the MgO adatom energy during the RF sputtering, as well as the thermal stability of the structure during the post-growth thermal annealing. Two different failure modes of the magnetoresistance are highlighted, involving with the decay of perpendicular magnetic anisotropy and destruction of coherent tunneling channels, respectively. Through the careful control of interfacial oxidation level and proper selection of the heavy metal layers, both perpendicular magnetic anisotropy and tunneling magnetoresistance of the junctions can be increased.

physics.app-ph↗

arXiv2022

Probabilistic Dalek -- Emulator framework with probabilistic prediction for supernova tomography

Supernova spectral time series can be used to reconstruct a spatially resolved explosion model known as supernova tomography. In addition to an observed spectral time series, a supernova tomography requires a radiative transfer model to perform the inverse problem with uncertainty quantification for a reconstruction. The smallest parametrizations of supernova tomography models are roughly a dozen parameters with a realistic one requiring more than 100. Realistic radiative transfer models require tens of CPU minutes for a single evaluation making the problem computationally intractable with traditional means requiring millions of MCMC samples for such a problem. A new method for accelerating simulations known as surrogate models or emulators using machine learning techniques offers a solution for such problems and a way to understand progenitors/explosions from spectral time series. There exist emulators for the TARDIS supernova radiative transfer code but they only perform well on simplistic low-dimensional models (roughly a dozen parameters) with a small number of applications for knowledge gain in the supernova field. In this work, we present a new emulator for the radiative transfer code TARDIS that not only outperforms existing emulators but also provides uncertainties in its prediction. It offers the foundation for a future active-learning-based machinery that will be able to emulate very high dimensional spaces of hundreds of parameters crucial for unraveling urgent questions in supernovae and related fields.

cs.LG↗