arXiv subjects

Arnau Quera-Bofarull

Publications and source records attributed to Arnau Quera-Bofarull.

13 recordsLinked to original sources

arXiv2026

$t_0$: A Time-Series Foundation Model for Forecasting with Context

We present $t_0$, a family of open-weights foundation models for forecasting with multivariate context. We release its first two members: $\texttt{t0-alpha}$ and $\texttt{t0-beta}$, respectively 102M and 256M parameters. Both condition their forecasts on target history, past covariates, and known-future covariates, without task-specific retraining. Their transformer layers alternate attention along time and across variates. They produce probabilistic forecasts through quantile predictions. Pretraining combines curated public data with synthetic generator families constructed to contain covariate-to-target dependencies. On GIFT-Eval, $\texttt{t0-alpha}$ reaches an aggregate CRPS of 0.4941, and $\texttt{t0-beta}$ a CRPS of 0.4738 and a MASE of 0.6865, third on both and within 4.0% of the best zero-shot TSFM. On fev-bench they score 42.2 and 46.7 in skill, the latter third again and 2.0 points behind the leader. We analyze $\texttt{t0-alpha}$ in depth. Known-future covariates raise its skill by 6.3 percentage points across 30 tasks. The report also examines its calibration, its rollout strategy on long horizons, and its robustness to missing data. On the Victoria electricity-demand benchmark, $\texttt{t0-beta}$ is among the most accurate models with a context of nearly a year. In an independent Macrocosm evaluation of hourly ERCOT prices over 29 months, both cut the MAE of the lagged-price baseline by 38%.

cs.LG↗

arXiv2026

Differentiable Electricity-Market Clearing for Gradient-Based Planning

Planning a large data center is difficult because a facility big enough to matter changes the electricity prices it will pay. Those prices are set by market clearing, a constrained optimization problem solved anew in every operating condition. However, simulating the market tells a planner how a candidate plan performs but not how to improve it. Here we treat market clearing as a differentiable optimization layer: each forward pass solves the market, and reverse-mode automatic differentiation propagates the planning cost back through the cleared prices to the plan. After validating these gradients against finite differences, we apply them to a concrete problem: allocating 50 MW of data-center load across six candidate buses in two synthetic networks, under a fixed cost per active site, evaluated over 36 operating states. Judged against exhaustive enumeration of all site combinations, gradient optimization recovers the continuous allocations almost exactly, with worst-case objective gaps of 2.3\% and 8.5\% of the cost difference between the best and worst single site. Its one systematic error is instructive: near the costs at which a site should close, the smooth relaxation of the discrete site count shrinks the site rather than closing it, so discrete transitions arrive late. Differentiable market clearing thus turns market-aware planning into a problem gradients can search.

cs.LG↗

arXiv2025

Automatic Differentiation of Agent-Based Models

Agent-based models (ABMs) simulate complex systems by capturing the bottom-up interactions of individual agents comprising the system. Many complex systems of interest, such as epidemics or financial markets, involve thousands or even millions of agents. Consequently, ABMs often become computationally demanding and rely on the calibration of numerous free parameters, which has significantly hindered their widespread adoption. In this paper, we demonstrate that automatic differentiation (AD) techniques can effectively alleviate these computational burdens. By applying AD to ABMs, the gradients of the simulator become readily available, greatly facilitating essential tasks such as calibration and sensitivity analysis. Specifically, we show how AD enables variational inference (VI) techniques for efficient parameter calibration. Our experiments demonstrate substantial performance improvements and computational savings using VI on three prominent ABMs: Axtell's model of firms; Sugarscape; and the SIR epidemiological model. Our approach thus significantly enhances the practicality and scalability of ABMs for studying complex systems.

cs.MA↗

arXiv2025

A differentiable model of supply-chain shocks

Modelling how shocks propagate in supply chains is an increasingly important challenge in economics. Its relevance has been highlighted in recent years by events such as Covid-19 and the Russian invasion of Ukraine. Agent-based models (ABMs) are a promising approach for this problem. However, calibrating them is hard. We show empirically that it is possible to achieve speed ups of over 3 orders of magnitude when calibrating ABMs of supply networks by running them on GPUs and using automatic differentiation, compared to non-differentiable baselines. This opens the door to scaling ABMs to model the whole global supply network.

physics.soc-ph↗

arXiv2024

On the limits of agency in agent-based models

Agent-based modeling (ABM) offers powerful insights into complex systems, but its practical utility has been limited by computational constraints and simplistic agent behaviors, especially when simulating large populations. Recent advancements in large language models (LLMs) could enhance ABMs with adaptive agents, but their integration into large-scale simulations remains challenging. This work introduces a novel methodology that bridges this gap by efficiently integrating LLMs into ABMs, enabling the simulation of millions of adaptive agents. We present LLM archetypes, a technique that balances behavioral complexity with computational efficiency, allowing for nuanced agent behavior in large-scale simulations. Our analysis explores the crucial trade-off between simulation scale and individual agent expressiveness, comparing different agent architectures ranging from simple heuristic-based agents to fully adaptive LLM-powered agents. We demonstrate the real-world applicability of our approach through a case study of the COVID-19 pandemic, simulating 8.4 million agents representing New York City and capturing the intricate interplay between health behaviors and economic outcomes. Our method significantly enhances ABM capabilities for predictive and counterfactual analyses, addressing limitations of historical data in policy design. By implementing these advances in an open-source framework, we facilitate the adoption of LLM archetypes across diverse ABM applications. Our results show that LLM archetypes can markedly improve the realism and utility of large-scale ABMs while maintaining computational feasibility, opening new avenues for modeling complex societal challenges and informing data-driven policy decisions.

cs.MA↗

arXiv2024

Private Agent-Based Modeling

The practical utility of agent-based models in decision-making relies on their capacity to accurately replicate populations while seamlessly integrating real-world data streams. Yet, the incorporation of such data poses significant challenges due to privacy concerns. To address this issue, we introduce a paradigm for private agent-based modeling wherein the simulation, calibration, and analysis of agent-based models can be achieved without centralizing the agents attributes or interactions. The key insight is to leverage techniques from secure multi-party computation to design protocols for decentralized computation in agent-based models. This ensures the confidentiality of the simulated agents without compromising on simulation accuracy. We showcase our protocols on a case study with an epidemiological simulation comprising over 150,000 agents. We believe this is a critical step towards deploying agent-based models to real-world applications.

cs.MA↗

arXiv2023

Some challenges of calibrating differentiable agent-based models

Agent-based models (ABMs) are a promising approach to modelling and reasoning about complex systems, yet their application in practice is impeded by their complexity, discrete nature, and the difficulty of performing parameter inference and optimisation tasks. This in turn has sparked interest in the construction of differentiable ABMs as a strategy for combatting these difficulties, yet a number of challenges remain. In this paper, we discuss and present experiments that highlight some of these challenges, along with potential solutions.

cs.MA↗

arXiv2023

Bayesian calibration of differentiable agent-based models

Agent-based modelling (ABMing) is a powerful and intuitive approach to modelling complex systems; however, the intractability of ABMs' likelihood functions and the non-differentiability of the mathematical operations comprising these models present a challenge to their use in the real world. These difficulties have in turn generated research on approximate Bayesian inference methods for ABMs and on constructing differentiable approximations to arbitrary ABMs, but little work has been directed towards designing approximate Bayesian inference techniques for the specific case of differentiable ABMs. In this work, we aim to address this gap and discuss how generalised variational inference procedures may be employed to provide misspecification-robust Bayesian parameter inferences for differentiable ABMs. We demonstrate with experiments on a differentiable ABM of the COVID-19 pandemic that our approach can result in accurate inferences, and discuss avenues for future work.

cs.MA↗

arXiv2023

Differentiable Agent-based Epidemiology

Mechanistic simulators are an indispensable tool for epidemiology to explore the behavior of complex, dynamic infections under varying conditions and navigate uncertain environments. Agent-based models (ABMs) are an increasingly popular simulation paradigm that can represent the heterogeneity of contact interactions with granular detail and agency of individual behavior. However, conventional ABM frameworks are not differentiable and present challenges in scalability; due to which it is non-trivial to connect them to auxiliary data sources. In this paper, we introduce GradABM: a scalable, differentiable design for agent-based modeling that is amenable to gradient-based learning with automatic differentiation. GradABM can quickly simulate million-size populations in few seconds on commodity hardware, integrate with deep neural networks and ingest heterogeneous data sources. This provides an array of practical benefits for calibration, forecasting, and evaluating policy interventions. We demonstrate the efficacy of GradABM via extensive experiments with real COVID-19 and influenza datasets.

cs.LG↗

arXiv2021

Qwind3: UV line-driven accretion disc wind models for AGN feedback

The ultraviolet (UV) bright accretion disc in active galactic nuclei (AGN) should give rise to line driving, producing a powerful wind which may play an important role in AGN feedback as well as in producing structures like the broad line region. However, coupled radiation-hydrodynamics codes are complex and expensive, so we calculate the winds instead using a non-hydrodynamical approach (the Qwind framework). The original Qwind model assumed the initial conditions in the wind, and had only simple radiation transport. Here, we present an improved version which derives the wind initial conditions and has significantly improved ray-tracing to calculate the wind absorption self consistently given the extended nature of the UV emission. We also correct the radiation flux for relativistic effects, and assess the impact of this on the wind velocity. These changes mean the model is more physical, so its predictions are more robust. We find that, even when accounting for relativistic effects, winds can regularly achieve velocities $\simeq$ (0.1-0.5) $c$, and carry mass loss rates which can be up to 30% of the accreted mass for black hole masses of $10^{7-9}$ $\mathrm{M}_\odot$, and mass accretion rates of 50% of the Eddington rate. Overall, the wind power scales as a power law with the black hole mass accretion rate, unlike the weaker scaling generally assumed in current cosmological simulations that include AGN feedback. The updated code, Qwind3, is publicly available in GitHub

astro-ph.HE↗

arXiv2020

Qwind code release: a non-hydrodynamical approach to modelling line-driven winds in active galactic nuclei

Ultraviolet (UV) line driven winds may be an important part of the active galactic nucleus (AGN) feedback process, but understanding their impact is hindered by the complex nature of the radiation hydrodynamics. Instead, we have taken the approach pioneered by Risaliti & Elvis, calculating only ballistic trajectories from radiation forces and gravity, but neglecting gas pressure. We have completely re-written their QWIND code using more robust algorithms, and can now quickly model the acceleration phase of these winds for any AGN spectral energy distribution spanning UV and X-ray wavebands. We demonstrate the code using an AGN with black hole mass $10^8\, M_\odot$ emitting at half the Eddington rate and show that this can effectively eject a wind with velocities $\simeq (0.1-0.2)\, c$. The mass loss rates can be up to $\simeq 0.3 M_\odot$ per year, consistent with more computationally expensive hydrodynamical simulations, though we highlight the importance of future improvements in radiation transfer along the multiple different lines of sight illuminating the wind. The code is fully public, and can be used to quickly explore the conditions under which AGN feedback can be dominated by accretion disc winds.

astro-ph.HE↗

arXiv2019

XNet: A convolutional neural network (CNN) implementation for medical X-Ray image segmentation suitable for small datasets

X-Ray image enhancement, along with many other medical image processing applications, requires the segmentation of images into bone, soft tissue, and open beam regions. We apply a machine learning approach to this problem, presenting an end-to-end solution which results in robust and efficient inference. Since medical institutions frequently do not have the resources to process and label the large quantity of X-Ray images usually needed for neural network training, we design an end-to-end solution for small datasets, while achieving state-of-the-art results. Our implementation produces an overall accuracy of 92%, F1 score of 0.92, and an AUC of 0.98, surpassing classical image processing techniques, such as clustering and entropy based methods, while improving upon the output of existing neural networks used for segmentation in non-medical contexts. The code used for this project is available online.

cs.CV↗

arXiv2018

Gravitational corrections to light propagation in a perturbed FLRW-universe and corresponding weak lensing spectra

When the gravitational lensing of the large-scale structure is calculated from a cosmological model a few assumptions enter: $(i)$ one assumes that the photons follow unperturbed background geodesics, which is usually referred to as the Born-approximation, $(ii)$ the lenses move slowly, $(iii)$ the source-redshift distribution is evaluated relative to the background quantities and $(iv)$ the lensing effect is linear in the gravitational potential. Even though these approximations are small individually they could sum up, especially since they include local effects such as the Sachs-Wolfe and peculiar motion, but also non-local ones like the Born-approximation and the integrated Sachs-Wolfe effect. In this work we will address all points mentioned and perturbatively calculate the effect on a tomographic cosmic shear power spectrum of each effect individually as well as all cross-correlations. Our findings show that each effect is at least 4 to 5 orders of magnitude below the leading order lensing signal. Finally we sum up all effects to estimate the overall impact on parameter estimation by a future cosmological weak lensing survey such as Euclid in a $w$CDM cosmology with parametrisation $Ω_\mathrm{m}$, $σ_8$,$n_\mathrm{s}$, $h$, $w_0$ and $w_\mathrm{a}$, using 5 tomographic bins. We consistently find a parameter bias of $10^{-5}$, which is therefore completely negligible for all practical purposes, confirming that other effects such as intrinsic alignments and magnification bias will be the dominant systematic source in future surveys.

astro-ph.CO↗