arXiv subjects

Tobias Huber

Publications and source records attributed to Tobias Huber.

At least 19 recordsLinked to original sources

arXiv2026

Update of the Standard-Model prediction for $\bar B \to X_s \gamma$

We report on two recent calculations on the inclusive radiative decay $\bar B\to X_s \gamma$, notably multi-parton contributions at NLO and the $Q_{1,2}-Q_7$ interference at NNLO for the physical value of the charm-quark mass. The former calculation formally completes $\bar{B} \rightarrow X_s \gamma$ at NLO in QCD at leading power, the latter removes a long-standing $\pm 3\%$ uncertainty arising from interpolation in $m_c$. The updated Standard-Model prediction for the CP- and isospin-averaged branching ratio reads ${\mathcal B}_{s \gamma}^{\rm SM} = (3.54 \pm 0.14)\times 10^{-4}$ for photon energies $E_\gamma > 1.6\,{\rm GeV}$ in the $B$-meson rest frame. This value is in good agreement with the current experimental average ${\mathcal B}_{s \gamma}^{\rm exp} = (3.49 \pm 0.19)\times 10^{-4}$.

hep-ph↗

arXiv2026

Three-loop QCD corrections to heavy-to-light form factors and applications to inclusive $B$ decays

We report on the calculation of heavy-to-light form factors at $\mathcal{O}(\alpha_s^3)$ and on selected phenomenological applications in inclusive $B$-decays. After outlining the loop calculation, we extract the hard function in $\bar B \to X_s \gamma$, and discuss our recent progress and preliminary results for the N$^3$LO corrections to partial decay rates in $\bar B \to X_u l \bar \nu_l$, important for the inclusive determination of $|V_{ub}|$. In particular, we establish relations for heavy-quark parameters in the shape-function scheme to four loops and improve a particular model of the $B$ meson shape-function.

hep-ph↗

arXiv2026

Preserve the Hard, Regenerate the Rest: Uncertainty-Guided Synthetic Training Data Augmentation with Diffusion Models

Semantic segmentation models struggle with data sparsity and rare or visually diverse regions, e.g., dense regions or small objects in aerial or autonomous mobility data. While synthetic augmentation is an appealing solution, directly generating new labeled data risks misalignment of labels and generated pixels. Existing solutions to this problem often rely on external models, or employ coarse heuristics such as indiscriminately augmenting all foreground objects or entire backgrounds, which wastes capacity on uninformative pixels. To address this, we propose an uncertainty-guided synthetic context augmentation strategy that strictly preserves label validity and efficiently maximizes pixel informativeness per synthetic sample - no external guardrails required. Using a baseline segmenter's predictive entropy, we identify uncertain semantic regions and inpaint only the complementary visual context. When fine-tuning the segmenter on this synthetic data, we compute the loss only over the original pixels, excluding inpainted regions. This focuses learning on the unmodified, uncertain regions while presenting them in novel contexts. We demonstrate substantial mIoU gains on Cityscapes, UAVID, and BDD100K with the largest gains on rare and difficult classes such as buses, trains, or (from the aerial perspective) cars. Our results demonstrate that uncertainty-guided context augmentation is a highly effective lever to improve segmentation performance on complex datasets, with code provided at https://github.com/XITASO/Preserve-the-Hard-Regenerate-the-Rest.

cs.CV↗

arXiv2026

QCD-factorization amplitudes from flavour symmetries: beyond the $SU(3)$ symmetric case

Using experimental information on branching ratios as well as direct and mixing-induced CP asymmetries, we perform a data-driven analysis of charmless non-leptonic $B \to PP$ decays, where $P$ is any of the light pseudoscalar mesons. Implementing flavour-$SU(3)$ breaking at the level of transition form factors, decay constants and phase space factors, we find a good fit to the current experimental data. Our best-fit point materializes in QCD-factorization amplitudes whose central values resemble many features of the dynamical predictions obtained within the QCD factorization framework. Moreover, we do not find any strong indications that the size of annihilation amplitudes is numerically enhanced beyond the na\"ive $\Lambda_{\textrm{QCD}}/m_b$ scaling. Subsequently, we address a number of phenomenological applications, among which are various flavour puzzles that have been persisting in non-leptonic $B$ decays for quite some time.

hep-ph↗

arXiv2026

SUNSET - A Sensor-fUsioN based semantic SegmEnTation exemplar for ROS-based self-adaptation

The fact that robots are getting deployed more often in dynamic environments, together with the increasing complexity of their software systems, raises the need for self-adaptive approaches. In these environments robotic software systems increasingly encounter (1) failures whose symptoms are easy to observe but root causes might be ambiguous or (2) multiple failures appearing concurrently. We present SUNSET, a ROS2-based exemplar that enables rigorous, repeatable evaluation of architecture-based self-adaptation in such conditions. It implements a sensor fusion semantic-segmentation pipeline driven by a trained Machine Learning (ML) model whose input preprocessing can be perturbed to induce realistic performance degradations. The exemplar exposes five observable failures, each of which can be caused by different faults and supports concurrent failures spanning self-healing and self-optimisation. SUNSET includes the segmentation pipeline, a trained ML model, fault-injection scripts, a baseline controller for further comparisons, and step-by-step integration and evaluation documentation to facilitate reproducible studies. The code is available at https://github.com/XITASO/sunset.

cs.RO↗

arXiv2025

Subatomic Heroes

Sharing the amazing achievements of the (particle) physics world with the general public is at the heart of the mission of the Subatomic Heroes, based at the University of Siegen, Germany. Originally this started out as an endeavor of theoretical particle physics, now we are steadily spreading out to cover and include more branches of physics and science. Our activities range from merging art with public physics lectures via marvelous artistic performances at the local theater, over dedicated events for high-school students, to our Subatomic Heroes channel on Instagram and TikTok where you may also find out when and where our famous "hadronic ice-cream" will be served next! So follow us on https://www.instagram.com/subatomic_heroes and https://www.tiktok.com/@subatomic_heroes.

physics.ed-ph↗

arXiv2025

Multi-parton contributions to $\bar B \to X_s \gamma$ at NLO

Many contributions to the decay rate of the inclusive radiative $\bar{B}\rightarrow X_s \gamma$ transition have been calculated to NNLO in QCD during the past decades. However, there are still a few unknown contributions from multi-parton final states which are formally NLO. In the present work, we compute those four-body $b \rightarrow s\, q\, \bar{q}\,\gamma$ contributions at NLO in QCD which need to be supplemented by the five-body $b \rightarrow s\, q\, \bar{q}\, g\,\gamma $ bremsstrahlung. This calculation formally completes the purely perturbative contributions to $\bar{B}\rightarrow X_s \gamma$ at NLO. Our results are obtained by applying modern techniques of integral reduction and evaluation of master integrals. In particular, the analytic integration over the four and five-particle phase space in the presence of a cut on the photon energy turns out to be technically involved. We give our results completely analytically in terms of multiple polylogarithms, including the dependence on the collinear logarithms which arise from the mass-regularisation of collinear divergences. The numerical impact of multi-parton corrections on the $\bar{B}\rightarrow X_s \gamma$ decay rate turns out to be small, owing to a partial cancellation between LO and NLO contributions.

hep-ph↗

arXiv2025

Who Is Responsible? Self-Adaptation Under Multiple Concurrent Failures With Unknown Faults in Complex Robotic Systems

Robotic systems increasingly operate in dynamic, unpredictable environments, where tightly coupled sensors and software modules increase the probability of a single failure cascading across components. Therefore, multiple strategies can be plausible to resolve the underlying fault. Most existing selfadaptive approaches that have been applied to robotics assume predefined one-to-one failure-to-adaptation mappings. We present a ROS2-based self-adaptation approach building upon MAPE-K that addresses (1) multiple simultaneous failures with differing criticality, (2) cascading failures across components, and (3) multiple plausible resolving strategies per detected failure. Central to our approach is an adaptation rule set which lets designers specify failure patterns, assign criticality levels, and enumerate multiple plausible adaptation strategies. This rule set, combined with an automatically extracted live dependency graph, enables lightweight root-cause analysis and strategy ranking to prioritize minimal and effective adaptations. Our approach implements a lightweight self-optimizing component which learns estimated success probabilities of different strategies for each known failure. Experiments on an underwater robot scenario and a perception use case show that our approach can identify root causes among concurrent failures, favors inexpensive adaptations, reduces unnecessary adaptations, and achieves performance comparable to existing baselines designed for sequential failures. The code is publicly available.

cs.RO↗

arXiv2024

Heavy-to-light form factors to three loops

We compute three-loop corrections of $\mathcal{O}(α_{s}^3)$ to form factors with one massive and one massless quark coupling to an external vector, axialvector, scalar, pseudoscalar, or tensor current. We obtain analytic results for the color-planar contributions, for the contributions of light-quark loops, and the contributions with two heavy-quark loops. For the computation of the remaining master integrals we use the "expand and match" approach which leads to semi-analytic results for the form factors. We implement our results in a {\tt Mathematica} and a {\tt Fortran} code which allows for fast and precise numerical evaluations in the physically relevant phase space. The form factors are used to compute the hard matching coefficients in Soft-Collinear Effective Theory for all currents. The tensor coefficients at light-like momentum transfer are used to extract the hard function in $\bar B \to X_s γ$ to three loops.

hep-ph↗

arXiv2024

Summary of CKM 2023 working group 5: Direct CP violation (DCPV) including $ϕ_{3}/γ$ from $B\to DK$, DCPV effects, branching fractions and polarisation in charmless $B_{(s)}$ decays

In this contribution a summary of the activities of Working Group 5 (WG5) presented during the 12th International Workshop on the CKM Unitarity Triangle (CKM2023) is reported. This includes new results on $ϕ_{3}/γ$ measurements using $B\to DK$ decays, search for $CP$ violation using charmless $B$ decays and $b$-Baryon decays, measurement of branching ratios in hadronic $B$ to charm decays, and theory of three-body nonleptonic $B$ decays.

hep-ex↗

arXiv2024

Relevant Irrelevance: Generating Alterfactual Explanations for Image Classifiers

In this paper, we demonstrate the feasibility of alterfactual explanations for black box image classifiers. Traditional explanation mechanisms from the field of Counterfactual Thinking are a widely-used paradigm for Explainable Artificial Intelligence (XAI), as they follow a natural way of reasoning that humans are familiar with. However, most common approaches from this field are based on communicating information about features or characteristics that are especially important for an AI's decision. However, to fully understand a decision, not only knowledge about relevant features is needed, but the awareness of irrelevant information also highly contributes to the creation of a user's mental model of an AI system. To this end, a novel approach for explaining AI systems called alterfactual explanations was recently proposed on a conceptual level. It is based on showing an alternative reality where irrelevant features of an AI's input are altered. By doing so, the user directly sees which input data characteristics can change arbitrarily without influencing the AI's decision. In this paper, we show for the first time that it is possible to apply this idea to black box models based on neural networks. To this end, we present a GAN-based approach to generate these alterfactual explanations for binary image classifiers. Further, we present a user study that gives interesting insights on how alterfactual explanations can complement counterfactual explanations.

cs.CV↗

arXiv2024

Inclusive $\bar{B}\to X_s \ell^+\ell^-$ at the LHC: theory predictions and new-physics reach

We present theoretical predictions for observables in inclusive $\bar{B}\to X_s \ell^+\ell^-$ suitable for measurements at hadron colliders through a sum-over-exclusive approach. At low $q^2$ we calculate the branching ratio and three angular observables. At high $q^2$ we provide the branching ratio and the ratio of the $\bar B \to X_s \ell^+\ell^-$ rate with respect to the inclusive $\bar B \to X_u \ell \bar\nu$ rate with the same phase-space cut. We compare our predictions to the $B$ factory measurements and also to an extraction of the experimental rate through a sum-over-exclusive method using branching ratios of the exclusive $\bar B \to K^{(*)}\mu^+\mu^-$ modes measured at LHCb. We find a consistent picture comparing Standard Model theory and experiment. As such, our analysis does not support a recent claim about a deficit in the inclusive branching ratio in the high-$q^2$ region. Finally, we present current model-independent bounds on new physics and emphasize the potential of complementary analyses of $\bar{B}\to X_s \ell^+\ell^-$ at Belle II and the LHC.

hep-ph↗

arXiv2023

Does Difficulty even Matter? Investigating Difficulty Adjustment and Practice Behavior in an Open-Ended Learning Task

Difficulty adjustment in practice exercises has been shown to be beneficial for learning. However, previous research has mostly investigated close-ended tasks, which do not offer the students multiple ways to reach a valid solution. Contrary to this, in order to learn in an open-ended learning task, students need to effectively explore the solution space as there are multiple ways to reach a solution. For this reason, the effects of difficulty adjustment could be different for open-ended tasks. To investigate this, as our first contribution, we compare different methods of difficulty adjustment in a user study conducted with 86 participants. Furthermore, as the practice behavior of the students is expected to influence how well the students learn, we additionally look at their practice behavior as a post-hoc analysis. Therefore, as a second contribution, we identify different types of practice behavior and how they link to students' learning outcomes and subjective evaluation measures as well as explore the influence the difficulty adjustment methods have on the practice behaviors. Our results suggest the usefulness of taking into account the practice behavior in addition to only using the practice performance to inform adaptive intervention and difficulty adjustment methods.

cs.HC↗

arXiv2023

Inclusive $\bar{B} \to X_s \ell^+ \ell^-$ with a hadronic mass cut

The hadronic mass spectrum of inclusive $\bar{B} \to X_s \ell^+ \ell^-$ is investigated at next to leading order in the heavy quark expansion. For mild cuts on the hadronic mass, the expansion, which applies when the cut is released, remains convergent. However, the cuts used at BaBar and Belle to reduce backgrounds from charged current semileptonic processes are too severe for a description in terms of matrix elements of local operators to apply. Strategies for interpolating between the two regions are discussed.

hep-ph↗

arXiv2023

Feynman integral reduction using Gröbner bases

We investigate the reduction of Feynman integrals to master integrals using Gröbner bases in a rational double-shift algebra Y in which the integration-by-parts (IBP) relations form a left ideal. The problem of reducing a given family of integrals to master integrals can then be solved once and for all by computing the Gröbner basis of the left ideal formed by the IBP relations. We demonstrate this explicitly for several examples. We introduce so-called first-order normal-form IBP relations which we obtain by reducing the shift operators in Y modulo the Gröbner basis of the left ideal of IBP relations. For more complicated cases, where the Gröbner basis is computationally expensive, we develop an ansatz based on linear algebra over a function field to obtain the normal-form IBP relations.

hep-ph↗

arXiv2023

GANterfactual-RL: Understanding Reinforcement Learning Agents' Strategies through Visual Counterfactual Explanations

Counterfactual explanations are a common tool to explain artificial intelligence models. For Reinforcement Learning (RL) agents, they answer "Why not?" or "What if?" questions by illustrating what minimal change to a state is needed such that an agent chooses a different action. Generating counterfactual explanations for RL agents with visual input is especially challenging because of their large state spaces and because their decisions are part of an overarching policy, which includes long-term decision-making. However, research focusing on counterfactual explanations, specifically for RL agents with visual input, is scarce and does not go beyond identifying defective agents. It is unclear whether counterfactual explanations are still helpful for more complex tasks like analyzing the learned strategies of different agents or choosing a fitting agent for a specific task. We propose a novel but simple method to generate counterfactual explanations for RL agents by formulating the problem as a domain transfer problem which allows the use of adversarial learning techniques like StarGAN. Our method is fully model-agnostic and we demonstrate that it outperforms the only previous method in several computational metrics. Furthermore, we show in a user study that our method performs best when analyzing which strategies different agents pursue.

cs.LG↗

arXiv2022

On the contribution of the electromagnetic dipole operator ${\cal O}_7$ to the $\bar B_s \to μ^+μ^-$ decay amplitude

We construct a factorization theorem that allows to systematically include QCD corrections to the contribution of the electromagnetic dipole operator in the effective weak Hamiltonian to the $\bar B_s \to μ^+μ^-$ decay amplitude. We first rederive the known result for the leading-order QED box diagram, which features a double-logarithmic enhancement associated to the different rapidities of the light quark in the $\bar B_s$ meson and the energetic muons in the final state. We provide a detailed analysis of the cancellation of the related endpoint divergences appearing in individual momentum regions, and show how the rapidity logarithms can be isolated by suitable subtractions applied to the corresponding bare factorization theorem. This allows us to include in a straightforward manner the QCD corrections arising from the renormalization-group running of the hard matching coefficient of the electromagnetic dipole operator in soft-collinear effective theory, the hard-collinear scattering kernel, and the $B_s$-meson distribution amplitude. Focusing on the contribution from the double endpoint logarithms, we derive a compact formula that resums the leading-logarithmic QCD corrections.

hep-ph↗

arXiv2022

Integrating Policy Summaries with Reward Decomposition for Explaining Reinforcement Learning Agents

Explaining the behavior of reinforcement learning agents operating in sequential decision-making settings is challenging, as their behavior is affected by a dynamic environment and delayed rewards. Methods that help users understand the behavior of such agents can roughly be divided into local explanations that analyze specific decisions of the agents and global explanations that convey the general strategy of the agents. In this work, we study a novel combination of local and global explanations for reinforcement learning agents. Specifically, we combine reward decomposition, a local explanation method that exposes which components of the reward function influenced a specific decision, and HIGHLIGHTS, a global explanation method that shows a summary of the agent's behavior in decisive states. We conducted two user studies to evaluate the integration of these explanation methods and their respective benefits. Our results show significant benefits for both methods. In general, we found that the local reward decomposition was more useful for identifying the agents' priorities. However, when there was only a minor difference between the agents' preferences, then the global information provided by HIGHLIGHTS additionally improved participants' understanding.

cs.LG↗