IRN Terascale @ Heidelberg
International Academic Forum - Heidelberg University
We are looking forward to welcoming you for the fall meeting of the IRN Terascale in Heidelberg on 19-21 October 2026.
A video connection link will be sent to the terascale general mailing list and to all registered participants.
For any questions regarding the corresponding session, please contact the group conveners:
More information at terascale.in2p3.fr
Monday evening will have a small reception and Tuesday's lunch will be provided.
Zoom link: https://cern.zoom.us/j/65784575146?pwd=y3RRN3aMJEeYRmRMipyEwHlTfPwQ4x.1
-
-
14:00
→
14:20
Welcome 20m
-
14:20
→
15:35
Dark Universe
-
14:20
Latest results from LZ 35m
The LUX-ZEPLIN experiment is a liquid-gas dual phase time projection chamber primarily searching for dark matter in the form of weakly interacting massive particles (WIMPs). It has been collecting science data with a 7-tonne active xenon target near-continuously since 2021. LZ has reported world-leading WIMP dark matter searches, a measurement of coherent solar $^8$B neutrino scattering on nuclei, and searches for other rare processes. Recently, LZ conducted a search for dark matter interactions using higher energy nuclear recoils than predicted in a conventional elastic WIMP search, observing an event at 248 keV, in tension with our current background-only hypothesis at 2.6 $\sigma$ global significance. This talk will introduce the signal models used to motivate and conduct high energy searches as well as on the backgrounds faced at these energies.
Orateur: Dr Joshua Green (University College London) -
14:55
Interpreting the LZ Excess: Models, Constraints and Phenomenological Challenges 40m
The recent high-energy recoil event reported by LUX-ZEPLIN has motivated a broad range of dark-matter interpretations. In this talk I will review the main theoretical scenarios that have been proposed, including thermal Higgsino and minimal electroweak dark matter, pseudo-Dirac models, endothermic and exothermic dark matter, and elastic momentum-dependent interactions.
I will focus in particular on the theoretical and phenomenological challenges faced by these interpretations. Beyond reproducing the LZ recoil spectrum, viable models must be consistent with the dark-matter relic abundance, Solar capture and neutrino-telescope constraints, collider searches, indirect detection, and the cosmological evolution of possible excited states. I will discuss the present status of the different scenarios and the complementary observations that can test them in the near future.
Orateur: MATTIA DI MAURO (INFN TORINO)
-
14:20
-
15:35
→
16:05
Coffee/Tea 30m
-
16:05
→
17:20
Dark Universe
-
16:05
Assessing the sensitivity to ALPs with VHE gamma-ray observations of selected AGN and galaxy cluster pairs 25m
Axion-Like Particles (ALPs) are well-motivated candidates for light dark matter and may couple with photons in the presence of astrophysical magnetic fields which give rise to the possibility of photon-ALP oscillations. If sufficiently strong, this mixing can induce characteristic energy-dependent feature in gamma-ray spectrum of astrophysical sources.
In this work, we explore the sensitivity of very-high-energy gamma-ray observations to ALP-photon interactions by combining measurements from AGNs located behind galaxy clusters. While the spectral features expected for a single source are highly sensitive to the specific magnetic-field configuration along the line of sight, averaging over a population of AGN-cluster pairs allows statistically robust predictions of the expected signal. To address this, we apply a stacking analysis of multiple AGN-cluster pairs, yielding a controlled prediction of the expected ALP induced spectral patterns and enhancing the sensitivity to such irregularities.
The method is evaluated using simulated observations of hard-spectrum Fermi-LAT AGNs accessible to Cherenkov instruments such as H.E.S.S. The combination of mock H.E.S.S. observations with our stacking approach enable exploration of the previously uncharted ALP dark matter parameter space in the neV mass range.
Orateur: cervane grimaud (IRFU / CEA Paris-Saclay) -
16:30
Light vector mediators as a window to dark matter 25m
The Standard Model (SM) of particle physics remains an incomplete theory, with the nature of dark matter (DM) among its major open questions. Guided by symmetry and simplicity, a new dark sector can communicate with the SM through three renormalizable portals: vector, scalar and neutrino. In particular, the vector portal is among the minimal realizations, with the standard dark photon as its benchmark for intensity-frontier and DM searches. To extract the full potential of light vector-mediated DM, in this talk we explore how we can relax the assumptions of the standard case in three directions: beyond the dark photon, by allowing for generic vector mediators, beyond the U(1), by considering non-abelian gauge structures, and beyond dark matter, by addressing other SM open problems, such as the origin of neutrino masses and the baryon asymmetry of the Universe. This approach highlights the potential of light vector-mediated DM models to serve as a rich and predictive window into physics beyond the Standard Model.
Orateur: Ana Foguel (University of Bologna) -
16:55
FIMPs in the Inert Doublet Model 25m
The Inert Doublet Model (IDM) is one of the most studied models of frozen-out WIMP dark matter. In this talk we will see that, depending on the thermal history of the universe, the IDM dark matter candidate can effectively behave as a FIMP despite being charged under the electroweak gauge group. We will discuss the DM abundance predicted by the IDM under different cosmological assumptions and some relevant phenomenological implications.
Orateur: Thomas REGGIO (LPCA)
-
16:05
-
17:20
→
18:20
Reception 1h
-
14:00
→
14:20
-
-
09:30
→
10:45
Beyond the Standard Model
-
09:30
An anomaly-free baryonic completion of the Standard Model 20m
We construct an explicit anomaly-free completion of a gauged baryon-number symmetry. The new gauge factor is $U(1)_{\tilde{B}}$: when restricted to the Standard Model fields its charge $\tilde{B}$ coincides with ordinary baryon number, whereas in the full theory it becomes an extended baryonic symmetry acting on both the Standard Model and a secluded sector. We show that anomaly cancellation alone leaves a broad class of solutions, but that requiring consistent symmetry breaking, viable decays of the exotic states, the absence of stable coloured relics, a neutral dark matter candidate protected by a $Z_2^{DM}$ parity, and improved high-scale gauge-coupling unification strongly constrains the structure of the model. These requirements necessitate enlarging the minimal secluded fermion sector by additional scalar fields and Yukawa interactions, giving rise to a light pseudoscalar that plays an important role in the dark matter phenomenology. We implement the resulting renormalisable model in SARAH and analyse its spectrum and phenomenology with SPheno, MicrOMEGAs, SModelS, and HiggsTools. We identify viable regions of parameter space containing a heavy baryonic Z', an even heavier secluded sector, a neutral dark matter candidate compatible with relic-density and direct-detection constraints, and improved gauge-coupling evolution relative to the Standard Model. In these regions, the relic abundance is determined either by co-annihilation with nearby secluded states or by annihilation into light pseudoscalars from the extended scalar sector.
Orateur: Arno Goudeau (LPTHE, Sorbonne University) -
09:55
Phenomenology of a $\{SU(2) \times U(1)\}^2$ Extension 20m
In this talk, we extend the electroweak gauge group of the Standard Model by introducing extra gauge factors, $SU(2)_{L}\times SU(2)_{H} \times U(1)_{1} \times U(1)_{2}$. We analyze theoretical consistency, Higgs constraints, electroweak precision tests, and dark matter direct detection bounds. Our results show that the model can accommodate viable dark matter scenarios while remaining compatible with current collider and astrophysical limits.
Orateur: Mohammed Khojali (IP2I - Université Lyon 1) -
10:20
Hunting for Heavy Neutral Leptons at Future Lepton Colliders 20m
Heavy neutral leptons offer a promising avenue to explore neutrino mass generation and physics beyond the Standard Model. In this talk, I will discuss their discovery prospects at future lepton colliders, focusing on cLFV scattering at μTRISTAN and enhanced di-Higgs production at high-energy lepton colliders. These complementary probes highlight the potential of future colliders to uncover the presence of heavy neutral leptons.
Orateur: Emanuelle Pinsard
-
09:30
-
10:45
→
11:15
Coffee/Tea 30m
-
11:15
→
12:30
Beyond the Standard Model
-
11:15
Probing Flavorful EFTs at the LHC 20m
Since the discovery of the Higgs boson, no clear signs of New Physics (NP) have been observed at the LHC. This absence is consistent with a separation between the scale of NP and the electroweak scale. In this scenario, Effective Field Theories (EFTs) provide a model-independent framework for analyzing LHC data and searching for indirect effects of physics beyond the Standard Model. In this talk, I will discuss how high-energy LHC processes, such as Drell–Yan and Higgs production channels, can probe NP scenarios relevant to semileptonic transitions, emphasizing the complementarity between high-energy LHC data and low-energy measurements of semileptonic and leptonic meson decays. In particular, I will illustrate this complementarity with two examples. The first shows how LHC data on Drell–Yan production can be used to test NP scenarios proposed to explain the mild discrepancies observed in recent BESIII data on $c\to s\mu\nu$ transitions. The second shows how associated Higgs production can constrain Wilson coefficients that are poorly constrained by Drell–Yan data, and how future projections for this channel can help clarify the current tensions in the extraction of the Cabibbo angle.
Orateur: Matheus Martines de Azevedo da Silva (University of São Paulo/Laboratoire De Physique Des 2 Infinite Irène Joliot-Curie) -
11:40
Vacuum Evolution of the R2HDM and the Induced Gravitational Wave and Collider Phenomenology 20m
Extended Higgs sectors beyond the Standard Model (BSM) allow to dynamically generate the observed baryon asymmetry of the Universe through electroweak baryogenesis and thereby solve one of the most prominent open problems of the SM. The strong first-order phase transitions (PTs), required to preserve the generated asymmetry in the electroweak vacuum, source gravitational waves (GW) that can be tested at future experiments like LISA. Gravitational waves hence provide the exciting possibility to probe BSM physics through cosmological processes. In order to be able to eventually pin down the specific underlying physics, a good understanding of the evolution of the vacuum of the model under investigation is indispensable as well as of the uncertainties that are involved in the derivation of the GW spectrum.
In this talk, the parameter space of the real, i.e. CP-conserving, 2-Higgs-Doublet Model is investigated with the public code BSMPTv3. Taking into account all relevant theoretical and experimental constraints, a thorough analysis of its vacuum evolution involving strong first-order phase transitions as well as the related collider and GW phenomenology is performed, complemented by an uncertainty discussion.
Orateur: Christoph Borschensky (Karlsruhe Institute of Technology) -
12:05
Modular symmetry models to explain leptonic flavour hierarchies 20m
The problem of hierarchy in flavour physics suggests the existence of a broken flavour symmetry. Modular symmetry is one of them but needs only one field, the modulus, which simultaneously plays the role of a flavon and Yukawa couplings are promoted to function of this modulus. Fields transformations under modular symmetry are characterized by an integer called the weight and a representation of a finite quotient group of the full modular group $SL(2,\mathbb{Z})$. It has been proven that modular symmetry can predict the masses and mixing angles of the fermionic sector of the Standard Model. Moreover, if the modulus takes special values, modular forms naturally acquire a hierarchy. This hierarchy depends on the choice of the finite quotient group and representations of the fields. We have investigated every possible choice and restrict them to a few number of representations able to naturally reproduce the hierarchies observed in the leptonic sector. Finally, we explicitly built one of these models based on the group $A_4\times\mathbb{Z}_4$ and found that with parameters of order one we can reproduce the masses and mixing angles of the leptonic sector.
Orateur: Victor Saulquin (Université de Strasbourg, IPHC)
-
11:15
-
12:30
→
14:00
Lunch 1h 30m
-
14:00
→
15:30
Methods and Tools
-
14:00
Anomaly detection - Review 25m
Despite strong motivation from theory and an intensive search
programme, no particle beyond the standard model has been
discovered at the LHC. Since model specific searches cannot
test every possible signal model, a broader approach capable
of testing many hypotheses at once is needed. Anomaly
detection refers to the set of machine-learning methods
specifically designed for this purpose.Starting from anomaly detection approaches used in industry
and the classical bump hunt, we will show how to construct
anomaly detection methods and review the first round of
experimental results.Orateur: Louis Moureaux (Universität Hamburg) -
14:30
Electroweak-inos in the pMSSM with SModelS v3.2 15m
In arXiv:2512.14502, we discussed the coverage of the electroweakino sector of the phenomenological minimal supersymmetric standard model (pMSSM) with SModelS v3.0, comparing with the ATLAS pMSSM study and investigating what can be gained by a global likelihood analysis. I will present an update of this study based on the recently released SModelS v3.2. The major novelties are the inclusion of mono-jet, disappearing-tracks and soft-leptons searches from ATLAS and/or CMS, and a much faster statistical evaluation in SModelS owing to machine-learned likelihoods.
Orateur: Leo Constantin (LPSC) -
14:50
Autoencoder-based anomaly detection and NAE 15m
Unsupervised anomaly detection with autoencoders is a promising data-driven and model-agnostic approach for new physics searches at the LHC. However, although neural network-based anomaly scores have transformative potential, they suffer from a lack of statistical interpretability. The normalized autoencoder (NAE) provides a probabilistic interpretation of the standard bottleneck architecture, tying the anomaly score to a learned likelihood. We validate this relation for a toy model, test it for jets using a dual-NAE setup, and show how a Bayesian NAE learns this likelihood with an uncertainty.
Orateur: Lorenz Vogel (ITP, Heidelberg University) -
15:10
The Information Geometry of Learned Latent Representations 15m
Many classification and reconstruction tasks in physics rely on learned latent representations of the data. When networks are trained with a notion of locality, they encode task-specific similarity as closeness in the latent space. Differential geometry, particularly information geometry, is a powerful tool to uncover the learned information in these latent representations and thereby retrace the decision-making of the network. With a jet tagging task in mind, we construct an information geometry on the learned latent space of a variational autoencoder with a classifier head. We show how the classifier likelihood links high-level physics observables to the learned latent geometry via the Fisher information metric. We supplement the Fisher information by advanced differential geometric concepts such as curvature and geodesic paths and use them to assess the importance of different features for the classification decision. We realize the importance of the likelihood’s skewness contribution as a geometric nonmetricity tensor and use it to construct coordinate-invariant measures of class separation.
Orateur: Rebecca Maria Kuntz (Astronomisches Rechen-Institut, Universität Heidelberg)
-
14:00
-
15:30
→
16:00
Coffee/Tea 30m
-
16:00
→
17:00
Methods and Tools
-
16:00
Neural Control Variates at LO and NLO 15m
We employ neural control variates to minimize the range of event weights and avoid negative weights for phase-space integration and event generation. A signed control variate, built from two normalizing flows, fulfills both tasks. Combined with neural importance sampling, it significantly reduces the computational cost of LO and NLO predictions. For the NLO case, our conditional neural control variate can be viewed as a trainable subtraction term, complementing the established physics subtraction schemes for enhanced sampling performance.
Orateur: Rebecca Revelli (Institute for theoretical physics Heidelberg) -
16:20
Sequential Simulation-Based Inference for the LHC 15m
Simulation-based inference (SBI) is a powerful tool for likelihood-free parameter estimation, but often requires large numbers of simulated events. At the LHC, where event generation can be computationally expensive, particularly with restrictive generator-level cuts and detector simulation, this can become a major bottleneck. We demonstrate sequential SBI for the inference of resonance masses and widths, showing that it can substantially improve sample efficiency by adaptively concentrating simulations in regions of parameter space favored by the data. This enables accurate inference with significantly fewer simulated events, making SBI more practical for computationally demanding LHC analyses.
Orateur: Ranit Das (Universität Heidelberg) -
16:40
MadAgents and Sfitter analyses 15m
We present an effective and communicative set of agents working with MadGraph and SFitter. Agentic installation, learning-by-doing training, and user support provide easy access to state-of-the-art simulations and accelerate LHC research. We show how MadAgents interact with inexperienced and advanced users, automatize event generation and autonomous simulation campaigns, and update a global SFitter analysis through agentic re-casting with a physicist in the loop. The agentic interface allows us to make advanced simulation and analysis methodologies available to a wider audience, with physical and technical aspects that can be generalized beyond MadGraph and SFitter.
Orateur: Daniel Schiller (Institute for Theoretical Physics, Heidelberg University)
-
16:00
-
17:15
→
18:00
Scientific Management Committee meeting (closed session) 45m
-
09:30
→
10:45
-
-
09:30
→
10:50
Higgs and Electroweak
-
09:30
Towards precise predictions of trilinear scalar couplings and di-Higgs production in extended scalar sectors 20m
Reconstructing the shape of the Higgs potential realised in Nature is crucial to understand the dynamics of the electroweak phase transition and probe physics beyond the Standard Model (BSM). For this purpose, precise theoretical predictions are needed for trilinear scalar couplings, which control the form of the potential, and for Higgs pair production cross-sections, which are the physical observables that allow accessing the trilinear couplings.
In this talk, I will present the public code framework anyBSM. Its first module, anyH3 allows automated complete one-loop computations of trilinear scalar couplings in arbitrary renormalisable theories, including momentum dependence and a high degree of flexibility in choices of renormalisation schemes. Additionally, I will highlight the new module anyHH which enables calculations of predictions di-Higgs production processes via gluon fusion at the (HL-)LHC, $gg\to h_ih_j$, in models with extended scalar sectors. I will present examples in multiple BSM scenarios, including some with resonant contributions, showing the impact of BSM corrections beyond the leading order. Finally, I will discuss recent works on corrections to trilinear scalar couplings at two loops and beyond, and their relevance for di-Higgs production.
Orateur: Dr Johannes Braathen (DESY) -
10:00
Impact of online b-Jet tagging on the ATLAS hh to bbtautau analysis 20m
A search for non-resonant hh production in the bbtt final state combines the full Run 2 dataset, corresponding to 140 fb^{-1} at √s = 13 TeV, with 56 fb^{-1} of early Run 3 data collected in 2022–2023 at √s = 13.6 TeV. Compared to the previous analysis in the same final state, the current analysis benefits from the additional integrated luminosity provided by the early Run 3 dataset, improved object reconstruction and identification algorithms, the introduction of online b-jet identification and its subsequent calibration, as well as several developments in the analysis strategy. This contribution focuses on the impact of online b-jet identification and its calibration on the analysis. Overall, these improvements result in an increased sensitivity to the hh signal. The measurement is compatible with the Standard Model expectation, with a best-fit signal strength of μₕₕ = 2.6^{+1.4}_{-1.0}. The observed (expected) significance of the hh signal relative to the background-only hypothesis reaches 2.6 (1.2) standard deviations, while the deviation from the Standard Model prediction amounts to 1.65 standard deviations. The same analysis is used to constrain the Higgs boson self-coupling. With the remaining Higgs boson couplings fixed to their Standard Model values, the observed (expected) 95% confidence-level interval is κ_λ ∈ [−3.4, 1.6] ∪ [5.5, 10.1] (κ_λ ∈ [−1.7, 8.5]). Finally, the analysis strategy is validated through measurements of Zh and ZZ production in the same final state. Their observed (expected) significances relative to the background-only hypothesis are 3.5 (2.4) and 1.8 (2.7) standard deviations, respectively.
Orateur: Christian Mavungu Tsava (CNRS) -
10:30
Longitudinal polarisation of electroweak gauge bosons in WZjj-EW production with ATLAS 20m
The scattering of longitudinally polarised electroweak gauge bosons provides a direct probe of the mechanism of electroweak symmetry breaking and of the structure of the Standard Model at high energies. This talk presents the first measurement of longitudinal polarisation states in electroweak $\mathrm{W}^{\pm}\mathrm{Z}$ production in association with two jets ($\mathrm{WZjj}$-EW), using 140 fb$^{-1}$ of ATLAS Run-2 data at $\sqrt{s}=13$ TeV and 164 fb$^{-1}$ of Run-3 data at $\sqrt{s}=13.6$ TeV.
The $\mathrm{WZjj}$-EW signal is isolated from backgrounds and the polarisation states are disentangled using deep neural networks, whose outputs are combined into an optimised two-dimensional discriminant for a binned maximum-likelihood fit. The fraction of events containing a longitudinally polarised $\mathrm{Z}$ boson is measured to be $f_{\mathrm{X0}} = 0.324^{+0.092}_{-0.085}$, with an observed (expected) significance of 4.0 (2.9) standard deviations, providing the first evidence for longitudinally polarised $\mathrm{Z}$ bosons in this process. Upper limits are also obtained on the fractions of events with a longitudinally polarised $\mathrm{W}$ boson and with both bosons longitudinally polarised. The $\mathrm{WZjj}$-EW fiducial cross-section at 13.6 TeV is measured for the first time and is found to be consistent with Standard Model predictions.
These measurements provide a new experimental probe of electroweak symmetry breaking through vector-boson scattering at the LHC.
Orateur: Mathis Dubau (LAPP)
-
09:30
-
11:00
→
11:20
Coffee / Tea 20m
-
11:20
→
11:50
Higgs and Electroweak
-
11:20
Top-Yukawa- and Light-Quark-Induced Electroweak Corrections to gg → HH 20m
After the discovery of the Higgs boson in 2012, the measurement of the Higgs self-coupling is still a challenge for current and future experiments in particle physics.
Higgs-boson pair production via gluon fusion is a loop-induced process. In order to increase the accuracy of the theoretical predictions for this process, higher-order corrections are necessary to reduce theoretical uncertainties and to describe differential distributions reliably. The next-to-leading order (NLO) corrections involve the evaluation of two-loop Feynman diagrams. In particular, for electroweak (EW) corrections, many different mass scales appear in the calculation, such as the gauge boson, bottom, top quark, and Higgs boson masses. Further complications include numerical instabilities due to virtual thresholds, which require careful treatment. In my talk, I will present results for the EW corrections induced by the top-Yukawa coupling with contributions from light-quark loops without using any reduction techniques to master integrals. The calculations are done by keeping the masses as fully symbolic parameters, allowing, in the future, for a study of parametric and mass scheme/scale uncertainties. I will furthermore discuss the next steps towards the calculation of the full EW corrections in our framework.Orateur: Sauro Carlotti (Karlsruhe Institute of Technology - ITP)
-
11:20
-
11:50
→
12:00
Famous last words 10m
-
09:30
→
10:50