BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Nicola Serra

Nicola Serra contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
hep-phhep-exastro-ph.COArtificial IntelligenceMachine Learningphysics.data-anphysics.ins-det

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
12works
0followers
7topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

12 published item(s)

preprint2026arXiv

Joint Treatment Effect Estimation from Incomplete Healthcare Data: Temporal Causal Normalizing Flows with LLM-driven Evolutionary MNAR Imputation

Target trial emulation (TTE) enables causal questions to be studied with observational data when randomized controlled trials (RCTs) are infeasible. Yet treatment-effect methods often address causal estimation, missingness, and temporal structure separately, limiting their robustness in electronic health records (EHRs), where time-varying confounding and missing-not-at-random (MNAR) biomarkers can reach 50%--80%. We propose a two-stage pipeline for treatment effect estimation from incomplete longitudinal EHRs. First, CausalFlow-T, a directed acyclic graph (DAG)-constrained normalizing flow with long short-term memory (LSTM)-encoded patient history, performs exact invertible counterfactual inference, avoiding approximation errors from variational inference and separating confounding through explicit causal structure. Ablations on four synthetic and one semi-synthetic benchmark with known counterfactuals show that DAG constraints and exact inference address distinct failure modes: neither compensates for the other. Second, because CausalFlow-T requires completed inputs, we introduce an LLM-driven evolutionary imputer that proposes executable imputation operators rather than individual entries, and evaluate it with three large language model (LLM) backends, including two open-source models. Across 30%--80% MNAR missingness, this imputer achieves the best pooled rank over biomarker and causal metrics, leading in point-wise accuracy and temporal extrapolation while preserving average treatment effect (ATE) recovery as statistical baselines degrade. On Swiss primary-care EHRs from adults with type 2 diabetes initiating a GLP-1 receptor agonist or SGLT-2 inhibitor, the pipeline estimates a per-protocol weight-loss difference of -0.98 kg [95% CI -1.01, -0.96] favoring GLP-1 receptor agonists, consistent with randomized evidence and obtained from realistically incomplete real-world EHRs.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Large Language Models for Physics Instrument Design

We study the use of large language models (LLMs) for physics instrument design and compare their performance to reinforcement learning (RL). Using only prompting, LLMs are given task constraints and summaries of prior high-scoring designs and propose complete detector configurations, which we evaluate with the same simulators and reward functions used in RL-based optimization. Although RL yields stronger final designs, we find that modern LLMs consistently generate valid, resource-aware, and physically meaningful configurations that draw on broad pretrained knowledge of detector design principles and particle--matter interactions, despite having no task-specific training. Based on this result, as a first step toward hybrid design workflows, we explore pairing the LLMs with a dedicated trust region optimizer, serving as a precursor to future pipelines in which LLMs propose and structure design hypotheses while RL performs reward-driven optimization. Based on these experiments, we argue that LLMs are well suited as meta-planners: they can design and orchestrate RL-based optimization studies, define search strategies, and coordinate multiple interacting components within a unified workflow. In doing so, they point toward automated, closed-loop instrument design in which much of the human effort required to structure and supervise optimization can be reduced.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Towards replacing detector simulation with heterogeneous GNNs in flavour physics analyses

Driven by the increasing volume of recorded data, the demand for simulation from experiments based at the Large Hadron Collider will rise sharply in the coming years. Addressing this demand solely with existing computationally intensive workflows is not feasible. This paper introduces a new fast simulation tool designed to address this demand at the LHCb experiment. This tool emulates the detector response to arbitrary multibody decay topologies at LHCb. Rather than memorising specific decay channels, the model learns generalisable patterns within the response, allowing it to interpolate to channels not present in the training data. Novel heterogeneous graph neural network architectures are employed that are designed to embed the physical characteristics of the task directly into the network structure. We demonstrate the performance of the tool across a range of decay topologies, showing the networks can correctly model the relationships between complex variables. The architectures and methods presented are generic and could readily be adapted to emulate workflows at other simulation-intensive particle physics experiments.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

A general effective field theory description of $b \to s l^+ l^-$ lepton universality ratios

We construct an expression for a general lepton flavour universality (LFU) ratio, $R_{X}$, in $b\to s l^+ l^-$ decays in terms of a series of hadronic quantities which can be treated as nuisance parameters. This expression allows to include any LFU ratio in global fits of $b\to s l^+ l^-$ short-distance parameters, even in the absence of a precise knowledge of the corresponding hadronic structure. The absence of sizeable LFU violation and the approximate left-handed structure of the Standard Model amplitude imply that only a very limited set of hadronic parameters hamper the sensitivity of $R_X$ to a possible LFU violation of short-distance origin. A global $b\to s l^+ l^-$ combination is performed including the measurement of $R_{pK}$ for the first time, resulting in a significance of new physics of $4.2\,σ$. In light of this, we evaluate the impact on the global significance of new physics using a set of experimentally promising non-exclusive $R_X$ measurements that LHCb can perform, and find that they can significantly increase the discovery potential of the experiment.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Probing effects of new physics in $Λ^0_{b}\toΛ^+_{c}μ^{-}\barν_μ$ decays

We present, for the first time, the six-fold differential decay density expression for $Λ^0_b\toΛ^+_{c} l^- \barν_{l}$, taking into account the polarisation of the $Λ^0_b$ baryon and a complete basis of new physics operators. Using the expected yield in the current dataset collected at the LHCb experiment, we present sensitivity studies to determine the experimental precision on the Wilson coefficients of the new physics operators with $Λ^0_{b}\toΛ^+_{c}μ^{-}\barν_μ$ decays in two scenarios. In the first case, unpolarised $Λ^0_{b}\toΛ^+_{c}μ^{-}\barν_μ$ decays with $Λ^+_c\to p K^+ π^-$ are considered, whereas polarised $Λ^0_{b}\toΛ^+_{c}μ^{-}\barν_μ$ decays with $Λ^+_c \to p K^0_S$ are studied in the second. For the latter scenario, the experimental precision that can be achieved on the determination of $Λ^0_b$ polarisation and $Λ^+_c$ weak decay asymmetry parameter is also presented.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

zfit: scalable pythonic fitting

Statistical modeling is a key element in many scientific fields and especially in High-Energy Physics (HEP) analysis. The standard framework to perform this task in HEP is the C++ ROOT/RooFit toolkit; with Python bindings that are only loosely integrated into the scientific Python ecosystem. In this paper, zfit, a new alternative to RooFit written in pure Python, is presented. Most of all, zfit provides a well defined high-level API and workflow for advanced model building and fitting, together with an implementation on top of TensorFlow, allowing a transparent usage of CPUs and GPUs. It is designed to be extendable in a very simple fashion, allowing the usage of cutting-edge developments from the scientific Python ecosystem in a transparent way. The main features of zfit are introduced, and its extension to data analysis, especially in the context of HEP experiments, is discussed.

0 citations0 reviewsNo code linkedOpen access
preprint2018arXiv

A global fit of the MSSM with GAMBIT

We study the seven-dimensional Minimal Supersymmetric Standard Model (MSSM7) with the new GAMBIT software framework, with all parameters defined at the weak scale. Our analysis significantly extends previous weak-scale, phenomenological MSSM fits, by adding more and newer experimental analyses, improving the accuracy and detail of theoretical predictions, including dominant uncertainties from the Standard Model, the Galactic dark matter halo and the quark content of the nucleon, and employing novel and highly-efficient statistical sampling methods to scan the parameter space. We find regions of the MSSM7 that exhibit co-annihilation of neutralinos with charginos, stops and sbottoms, as well as models that undergo resonant annihilation via both light and heavy Higgs funnels. We find high-likelihood models with light charginos, stops and sbottoms that have the potential to be within the future reach of the LHC. Large parts of our preferred parameter regions will also be accessible to the next generation of direct and indirect dark matter searches, making prospects for discovery in the near future rather good.

0 citations0 reviewsNo code linkedOpen access
preprint2018arXiv

Global fits of GUT-scale SUSY models with GAMBIT

We present the most comprehensive global fits to date of three supersymmetric models motivated by grand unification: the Constrained Minimal Supersymmetric Standard Model (CMSSM), and its Non-Universal Higgs Mass generalisations NUHM1 and NUHM2. We include likelihoods from a number of direct and indirect dark matter searches, a large collection of electroweak precision and flavour observables, direct searches for supersymmetry at LEP and Runs I and II of the LHC, and constraints from Higgs observables. Our analysis improves on existing results not only in terms of the number of included observables, but also in the level of detail with which we treat them, our sampling techniques for scanning the parameter space, and our treatment of nuisance parameters. We show that stau co-annihilation is now ruled out in the CMSSM at more than 95\% confidence. Stop co-annihilation turns out to be one of the most promising mechanisms for achieving an appropriate relic density of dark matter in all three models, whilst avoiding all other constraints. We find high-likelihood regions of parameter space featuring light stops and charginos, making them potentially detectable in the near future at the LHC. We also show that tonne-scale direct detection will play a largely complementary role, probing large parts of the remaining viable parameter space, including essentially all models with multi-TeV neutralinos.

0 citations0 reviewsNo code linkedOpen access
preprint2018arXiv

Prospects for disentangling long- and short-distance effects in the decays $B\to K^* μ^+μ^-$

Theory uncertainties on non-local hadronic effects limit the New Physics discovery potential of the rare decays $B\to K^*μ^+μ^-$. We investigate prospects to disentangle New Physics effects in the short-distance coefficients from these effects. Our approach makes use of an event-by-event amplitude analysis, and relies on the state of the art parametrisation of the non-local contributions. We find that non-standard effects in the short-distance coefficients can be successfully disentangled from non-local hadronic effects. The impact of the truncation on the parametrisation of non-local contributions to the Wilson coefficients are for the first time systematically examined and prospects for its precise determination are discussed. We find that physical observables are unaffected by these uncertainties. Compared to other methods, our approach provides for a more precise extraction of the angular observables from data.

0 citations0 reviewsNo code linkedOpen access
preprint2018arXiv

Status of the scalar singlet dark matter model

One of the simplest viable models for dark matter is an additional neutral scalar, stabilised by a $\mathbb{Z}_2$ symmetry. Using the GAMBIT package and combining results from four independent samplers, we present Bayesian and frequentist global fits of this model. We vary the singlet mass and coupling along with 13 nuisance parameters, including nuclear uncertainties relevant for direct detection, the local dark matter density, and selected quark masses and couplings. We include the dark matter relic density measured by Planck, direct searches with LUX, PandaX, SuperCDMS and XENON100, limits on invisible Higgs decays from the Large Hadron Collider, searches for high-energy neutrinos from dark matter annihilation in the Sun with IceCube, and searches for gamma rays from annihilation in dwarf galaxies with the Fermi-LAT. Viable solutions remain at couplings of order unity, for singlet masses between the Higgs mass and about 300 GeV, and at masses above $\sim$1 TeV. Only in the latter case can the scalar singlet constitute all of dark matter. Frequentist analysis shows that the low-mass resonance region, where the singlet is about half the mass of the Higgs, can also account for all of dark matter, and remains viable. However, Bayesian considerations show this region to be rather fine-tuned.

0 citations0 reviewsNo code linkedOpen access
preprint2017arXiv

FlavBit: A GAMBIT module for computing flavour observables and likelihoods

Flavour physics observables are excellent probes of new physics up to very high energy scales. Here we present FlavBit, the dedicated flavour physics module of the global-fitting package GAMBIT. FlavBit includes custom implementations of various likelihood routines for a wide range of flavour observables, including detailed uncertainties and correlations associated with LHCb measurements of rare, leptonic and semileptonic decays of B and D mesons, kaons and pions. It provides a generalised interface to external theory codes such as SuperIso, allowing users to calculate flavour observables in and beyond the Standard Model, and then test them in detail against all relevant experimental data. We describe FlavBit and its constituent physics in some detail, then give examples from supersymmetry and effective field theory illustrating how it can be used both as a standalone library for flavour physics, and within GAMBIT.

0 citations0 reviewsNo code linkedOpen access
preprint2017arXiv

GAMBIT: The Global and Modular Beyond-the-Standard-Model Inference Tool

We describe the open-source global fitting package GAMBIT: the Global And Modular Beyond-the-Standard-Model Inference Tool. GAMBIT combines extensive calculations of observables and likelihoods in particle and astroparticle physics with a hierarchical model database, advanced tools for automatically building analyses of essentially any model, a flexible and powerful system for interfacing to external codes, a suite of different statistical methods and parameter scanning algorithms, and a host of other utilities designed to make scans faster, safer and more easily-extendible than in the past. Here we give a detailed description of the framework, its design and motivation, and the current models and other specific components presently implemented in GAMBIT. Accompanying papers deal with individual modules and present first GAMBIT results. GAMBIT can be downloaded from gambit.hepforge.org.

0 citations0 reviewsNo code linkedOpen access