Searcharxiv⌕ Search

arXiv subjects

João Gonçalves

Publications and source records attributed to João Gonçalves.

At least 19 recordsLinked to original sources

Prominence: A discriminator of gravitational wave signals

The concept of prominence is familiar to signal engineers, topographers and mountaineers. We introduce Prominence $\cal P$ as a discriminator of gravitational wave (GW) signals. We treat black hole and neutron star binaries as astrophysical background sources, and show how $\cal P$ can be used to distinguish between GW spectra produced by first-order phase transitions, domain walls and cosmic strings, and combinations thereof. Prominence can also be used to discriminate between these and off-piste sources of GWs. The uncertainty in the measured energy density in GWs at Pulsar Timing Arrays needs to be at the sub-percent to percent level for $\cal{P}$ to achieve discrimination at 3$σ$. LISA and ET data are expected to have sufficiently small uncertainties that Prominence can play a central role in their analysis. We define and apply a $χ^2$ statistic based on an inner product in signal space to show that sources indistinguishable at $2σ$ using the signal-to-noise ratio can be distinguished using $\mathcal{P}$ with significances above and in some cases far above $3σ$.

hep-ph↗

Prior-matched evaluation of operational Earth-observation classifiers: a three-number reporting method demonstrated on Sentinel-1 internal-wave detection

The Internal Waves Service screens the Sentinel-1 Wave-mode archive for internal solitary waves, routing detections to experts whose adjudication time is the resource the effort exists to conserve. Because attention is the cost of error, precision leads. Its classifier was trained and reported at a one-to-one class balance, fixed before the operational rate could be known. That rate has since emerged at roughly one scene in twenty, and a balanced-test score badly overstates the precision a validator meets. A model that scores 0.794 balanced-test precision scores 0.192 in real operation: the gap is a systematic artefact of reporting at the wrong prior, invisible to the metric most work quotes. We show the mismatch to be an evaluation problem in the costume of a training one at a fixed recall, prior correction and calibration cannot move precision, and answer it with a prior-matched reporting method based on three numbers: balanced-test, operational-prior, and real post-deployment, whose contrast is the honest measure. A precision-first, leakage-controlled development cycle then improves the classifier lever by lever, each promoted only against a pre-registered margin; negative variety and the aggregation head lifting, capacity paying once then stopping, calibration inert, so the honest negatives are as much a result as the gains. Holding recall at a floor of 0.80 and certifying against a sealed, single-read lockbox, the promoted model reports 0.927 precision at the operational prior; an out-of-time check confirms discrimination transfers to unseen periods while a fixed operating point does not. Prior-matched reporting, begin balanced, then move to the prior as the stream reveals it, transfers to any operational Earth-observation service bootstrapping a rare-event detector under a prior it has yet to discover.

cs.LG↗

Pirateship: Append-Only Ledgers for (Mostly) Trusted Execution Environments

Distributed ledgers are increasingly relied upon by industry to provide trustworthy accountability, strong integrity protection, and high availability for critical data without centralizing trust. Recently, distributed append-only logs are opting for a layered approach, combining crash-fault-tolerant (CFT) consensus with hardware-based Trusted Execution Environments (TEEs) for greater resiliency. Unfortunately, hardware TEEs can be subject to (rare) attacks, undermining the very guarantees that distributed ledgers are carefully designed to achieve. In response, we present Pirateship, a new distributed consensus protocol that cautiously trusts the guarantees of TEEs. Pirateship carefully embeds a Byzantine fault-tolerant (BFT) protocol inside of a CFT protocol with no additional messages. This is made possible through careful refactoring of both the CFT and BFT protocols such that their structure aligns. Pirateship achieves performance in line with regular TEE-enabled consensus protocols, while guaranteeing integrity in the face of TEE platform compromises.

cs.DC↗

Design and Empirical Evaluation of a Network-Centric, On-Premises Architecture for Earth Observation Data Access

Earth observation (EO) programmes generate data at volumes that exceed the transfer and storage capacity of most institutional networks. Public cloud platforms address this for well-resourced organisations, but institutions across the Atlantic basin face constraints in connectivity, sovereignty and funding that make on-premises infrastructure the only viable path. Cloud-native data formats enable efficient partial reads, yet their performance depends on the bandwidth of the underlying network fabric, a dependency rarely measured in isolation. This paper presents a replicable, network-centric architecture for on-premises EO data access, evaluated at its first operational deployment: the AIR Data Centre, founding node of the Atlantic Cloud. The system comprises a MinIO object storage cluster on a 100 GbE fabric, a PostGIS metadata catalogue and an OGC API-EDR access layer. We characterise the fabric under sustained parallel load, evaluate object storage throughput for EO-representative workloads, and compare measured performance against throttled baselines on identical hardware, isolating network bandwidth as the sole variable. Multi-site replication benchmarks with partner institutions characterise the federation primitive the model depends on. Network bandwidth is the dominant constraint on storage throughput for bulk EO data access up to a threshold; beyond it, endpoint memory topology rather than capacity governs how much bandwidth a system can use. For this hardware class that threshold lies above 10 Gbps per server. Below it, network capacity alone sets what the facility can deliver; above it, the return on further network investment depends on endpoint memory provisioning, which can be deferred and bought later.

cs.DC↗

Gravitational Waves from Dark Gauge Sectors

We explore gravitational-wave (GW) signatures from a strong first-order phase transition in a non-Abelian dark sector, which naturally gives rise to vector dark matter (DM). We consider a general class of models featuring a new dark gauge sector communicating with the Standard Model (SM) through a Higgs portal and a vector-like fermionic portal. We also study the scenario where the dark sector interacts with the SM only via gravity. In all cases, we scan the full parameter space and analyse GW production and highlight the regions with visible GW signatures. Notably, the fermionic portal yields distinctive GW signals at LISA with peak frequencies of 1--10 mHz, reaching up to 1 Hz for future interferometers like BBO and DECIGO, while the Higgs portal scenario remains limited to around 1 mHz. Both frameworks account for the observed DM abundance and predict detectable LISA signals for dark vector bosons near 1--4 TeV, with a $\sim$10 GeV dark Higgs. Finally, we identify a unique six-top final state from pair-produced vector-like fermions, offering a striking collider signature within HL-LHC reach. Its detection would provide a smoking-gun signal for the fermionic portal, establishing complementarity between collider, GW, and DM signals.

hep-ph↗

Primordial black holes and magnetic fields in conformal neutrino mass models

Sufficiently strong and long-lasting first-order phase transitions can produce primordial black holes (PBHs) that contribute substantially to the dark matter abundance of the Universe, and can produce large-scale primordial magnetic fields. We study these mechanisms in a generic class of conformal $\mathrm{U(1)}^\prime$ models that also explain active neutrino oscillation data via the type-I seesaw mechanism. We find that phase transitions that occur at seesaw scales between $10^4$ GeV and $10^{11}$ GeV produce gravitational wave signals (from the dynamics of the phase transition and from the decay of cosmic string loops) at LISA/ET that can be correlated with microlensing signals of PBHs at the Roman Space Telescope, while scales near $10^{11}$ GeV can be correlated with Hawking evaporation signals at future gamma-ray telescopes. LISA can probe the entire range of PBH masses between $1\times 10^{-16}M_\odot$ and $8\times 10^{-11}M_\odot$ if PBHs fully account for the dark matter abundance. For Z' masses between 5 TeV and 100 TeV, and $\sim 3$ TeV right-handed neutrinos, helical magnetic fields can be produced with magnitudes $\sim 10^{-16}$-$10^{-13}$ G and coherence lengths $\sim 10^{-4}$-$10^{-2}$ Mpc, above current blazar lower bounds.

hep-ph↗

SHARE: Social-Humanities AI for Research and Education

This intermediate technical report introduces the SHARE family of base models and the MIRROR user interface. The SHARE models are the first causal language models fully pretrained by and for the social sciences and humanities (SSH). Their performance in modelling SSH texts is close to that of general purpose models (Phi-4) which use 100 times more tokens, as shown by our custom SSH Cloze benchmark. The MIRROR user interface is designed for reviewing text inputs from the SSH disciplines while preserving critical engagement. By prototyping a generative AI interface that does not generate any text, we propose a way to harness the capabilities of the SHARE models without compromising the integrity of SSH principles and norms.

cs.CL↗

Bentō: Optimizing Persistent Memory Programs

Persistent Memory (PM) is a new storage technology thatbrings high performance, byte addressability, and persistency for a lesser cost than DRAM. Due to cache volatility and store reordering, developers must use explicit instructions (e.g.: flush and fence) to guarantee that the application state remains consistent upon crashes. This is difficult to get right and, in fact, several tools have been created to detect bugs in PM programs. To overcome this difficulty, programmers tend to be overly conservative, for instance, by enforcing unnecessary ordering constraints, which partially forfeits the performance benefits of using PM. In this paper, we study the impact that different combinations of persistency instructions have in several PM programs and found that a specific combination can lead to performance improvements while preserving the original crash-consistency semantics. Based on these results we developed Bentō an automatic and black-box binary rewriter that can boost the performance of existing PM programs by up to 15% with minimal programmer effort.

cs.ET↗

Sterile neutrino dark matter in conformal Majoron models

We study sterile neutrino dark matter (DM) in a classically conformal U(1)' extension of the Standard Model with three right-handed neutrinos and a Majoron-like singlet scalar that generate the observed pattern of active neutrino masses and mixing via the type-I seesaw mechanism. Working in the regime of strongly suppressed active-sterile mixing, we show that the observed DM abundance can be produced through freeze-in from feeble interactions mediated by the heavy Z' and the conformal scalar. We solve the Boltzmann equation for the nonthermal phase-space distribution and confront the scenario with Lyman-$α$ data by computing the matter power spectrum. For keV-scale sterile neutrinos we identify the viable parameter space consistent with structure-formation and X-ray bounds, including regions compatible with a tentative 3.5 keV line. If a second sterile state is long-lived, late decays can realize a two-component setup that alleviates the $S_8$ tension. In a highly fine-tuned variant of the model, the 220 PeV KM3NeT event can also be explained by invoking the decay of a superheavy sterile neutrino.

hep-ph↗

Supercooled phase transitions in conformal dark sectors explain NANOGrav data

According to recent lore, it is difficult to explain the evidence for a stochastic gravitational wave background obtained by pulsar timing arrays with supercooled first-order phase transitions (FOPTs). We demonstrate that supercooled FOPTs in dark U(1)' models with a conformal dark sector easily explain the nHz signal at NANOGrav.

hep-ph↗

Rethinking PM Crash Consistency in the CXL Era

Persistent Memory (PM) introduces new opportunities for designing crash-consistent applications without the traditional storage overheads. However, ensuring crash consistency in PM demands intricate knowledge of CPU, cache, and memory interactions. Hardware and software mechanisms have been proposed to ease this burden, but neither proved sufficient, prompting a variety of bug detection tools. With the sunset of Intel Optane comes the rise of Compute Express Link (CXL) for PM. In this position paper, we discuss the impact of CXL's disaggregated and heterogeneous nature in the development of crash-consistent PM applications, and outline three research directions: hardware primitives, persistency frameworks, and bug detection tools.

cs.ET↗

The advantages of context specific language models: the case of the Erasmian Language Model

The current trend to improve language model performance seems to be based on scaling up with the number of parameters (e.g. the state of the art GPT4 model has approximately 1.7 trillion parameters) or the amount of training data fed into the model. However this comes at significant costs in terms of computational resources and energy costs that compromise the sustainability of AI solutions, as well as risk relating to privacy and misuse. In this paper we present the Erasmian Language Model (ELM) a small context specific, 900 million parameter model, pre-trained and fine-tuned by and for Erasmus University Rotterdam. We show how the model performs adequately in a classroom context for essay writing, and how it achieves superior performance in subjects that are part of its context. This has implications for a wide range of institutions and organizations, showing that context specific language models may be a viable alternative for resource constrained, privacy sensitive use cases.

cs.CL↗

Gravitational waves from supercooled phase transitions in conformal Majoron models of neutrino mass

We study supercooled first-order phase transitions above the QCD scale in a wide class of conformal Majoron-like U(1)' models that explain the totality of active neutrino oscillation data and produce a detectable stochastic gravitational wave background (SGWB) at LIGO, LISA and ET. We place constraints on the U(1)' breaking scale and gauge coupling using current LIGO-Virgo-Kagra data. We find that strong supercooling can be ruled out in large regions of parameter space if a SGWB is not detected by these experiments. A null signal at LIGO and ET will disfavor a type-I seesaw scale above $10^{14}$ GeV, while a positive signal is a signature of heavy right-handed neutrinos. On the other hand, LISA will be sensitive to seesaw scales as low as a TeV, and could detect a SGWB even if the right-handed neutrinos are decoupled.

hep-ph↗

Tridiagonal scalar mass matrix in the CP4 3HDM and its implications

When parametrizing multi-Higgs potentials, it is desirable to express its coefficients via observables. This is routinely done for the 2HDM, but this approach often fails in more elaborate models. Here, we show that the scalar sector of the CP4 3HDM, an intriguing model based on an order-4 $CP$ symmetry, can also be parametrized in an observable-driven manner. The key feature that makes it work is the very special tridiagonal form of the $5\times 5$ neutral Higgs mass matrix. We propose a set of input observables and present an algorithm to reconstruct the coefficients of the potential through linear relations. Equipped with this procedure, we explore the scalar sector of the CP4 3HDM beyond the limitations of previous studies. In particular, we identify a viable and testable regime in which all additional Higgses lie in the 300-600 GeV range. This work offers a key ingredient for a future full phenomenological scan of this model.

hep-ph↗

Combining Autoregressive and Autoencoder Language Models for Text Classification

This paper presents CAALM-TC (Combining Autoregressive and Autoencoder Language Models for Text Classification), a novel method that enhances text classification by integrating autoregressive and autoencoder language models. Autoregressive large language models such as Open AI's GPT, Meta's Llama or Microsoft's Phi offer promising prospects for content analysis practitioners, but they generally underperform supervised BERT based models for text classification. CAALM leverages autoregressive models to generate contextual information based on input texts, which is then combined with the original text and fed into an autoencoder model for classification. This hybrid approach capitalizes on the extensive contextual knowledge of autoregressive models and the efficient classification capabilities of autoencoders. Experimental results on four benchmark datasets demonstrate that CAALM consistently outperforms existing methods, particularly in tasks with smaller datasets and more abstract classification objectives. The findings indicate that CAALM offers a scalable and effective solution for automated content analysis in social science research that minimizes sample size requirements.

cs.CL↗

Gravitational waves from a scotogenic two-loop neutrino mass model

We propose a framework to account for neutrino masses at the two-loop level. This mechanism introduces new scalars and Majorana fermions to the Standard Model. It is assumed the existence of a global $\mathrm{U(1)\times \mathcal{Z}_2}$ symmetry which after partial breaking provides the stability of the dark matter candidates of the theory. The rich structure of the potential allows for the possibility of first-order phase transitions (FOPTs) in the early Universe which can lead to the generation of primordial gravitational waves as one of the potentially observable signatures of this model. Taking into account relevant constraints from lepton flavour violation, neutrino physics as well as the trilinear Higgs couplings at next-to-leading order accuracy, we have found a wide range of possible FOPTs which are strong enough to be probed at the proposed gravitational-wave interferometer experiments such as LISA.

hep-ph↗

On interplay between flavour anomalies and neutrino properties

A minimal extension of the Standard Model (SM) featuring two scalar leptoquarks, an SU(2) doublet with hypercharge 1/6 and a singlet with hypercharge 1/3, is proposed as an economical benchmark model for studies of an interplay between flavour physics and properties of the neutrino sector. The presence of such type of leptoquarks radiatively generates neutrino masses and offers a simultaneous explanation for the current B-physics anomalies involving $b \to c \ell ν_\ell$ decays. The model can also accommodate both the muon magnetic moment and the recently reported $W$ mass anomalies, while complying with the most stringent lepton flavour violating observables.

hep-ph↗

Exploring mixed lepton-quark interactions in non-resonant leptoquark production at the LHC

Searches for new physics (NP) at particle colliders typically involve multivariate analysis of kinematic distributions of final state particles produced in a decay of a hypothetical NP resonance. Since the pair-production cross-sections mediated by such resonances are strongly suppressed by the NP scale, this analysis becomes less relevant for NP searches for masses of the BSM resonance above 1 TeV. On the other hand, $t$-channel processes are less sensitive to the mass of the virtual mediator and therefore larger phase-space can be potentially probed as well as the couplings between the NP particles and the Standard Model fields. The fact that transitions between different generations of quarks and leptons may exist, the potential of the search presented in this article can be used, as a reference guide, to enlarge significantly the scope of searches performed at the LHC to flavour off-diagonal channels, in a theoretically consistent approach. In this work, we study non-resonant production of scalar leptoquarks which have been proposed in the literature to provide a potential avenue for radiative generation of neutrino masses, accommodating as well the existing flavour physics data. Final states involving just two muons at the LHC ($μ^+, μ^-$), are used as a well-motivated case study.

hep-ph↗