Search

The impact of cognitive biases on decision-making in the digital world remains under-explored despite its well-documented effects in physical contexts. This study addresses this gap by investigating the attractiveness halo effect using AI-based beauty filters. We conduct a large-scale online user study involving 2,748 participants who rated facial images from a diverse set of 462 distinct individuals in two conditions: original and attractive after applying a beauty filter. Our study reveals that the same individuals receive statistically significantly higher ratings of attractiveness and other traits, such as intelligence and trustworthiness, in the attractive condition. We also study the impact of age, gender, and ethnicity and identify a weakening of the halo effect in the beautified condition, resolving conflicting findings from the literature and suggesting that filters could mitigate this cognitive bias. Finally, our findings raise ethical concerns regarding the use of beauty filters., Comment: 40 pages, 15 figures, 13 tables; Version 2 incorporates feedback from the reviews and the format has been updated to match the requirements of the Royal Society Open Science

We examine the contribution of tensor modes, in addition to the dominant scalar ones, on the temperature anisotropies of the cosmic microwave background (CMB). To this end, we analyze in detail the temperature two-point angular correlation function $C(\theta)$ from the Planck 2018 dataset, focusing on large angles ($\theta \gtrsim 120^{\circ}$) corresponding to small $\ell$ multipoles. A hierarchical set of infrared cutoffs are naturally introduced to the scalar and tensor power spectra of the CMB by invoking an extra Kaluza-Klein dimension compactifying at about the GUT scale between the Planck epoch and the start of inflation. We associate this set of lower scalar and tensor cutoffs with the parity of the multipole expansion of the $C(\theta)$ function. By fitting the Planck 2018 data we compute the multipole coefficients thereby reproducing the well-known odd-parity preference in angular correlations seen by all three satellite missions COBE, WMAP and Planck. Our fits improve significantly once tensor modes are included in the analysis, hence providing a hint of the imprints of primordial gravitational waves on the temperature correlations observed in the CMB today. To conclude we suggest a relationship between, on the one hand, the lack of (positive) large-angle correlations and the odd-parity dominance in the CMB and, on the other hand, the effect of primordial gravitational waves on the CMB temperature anisotropies., Comment: 12 pages, 2 figures

Federated Learning (FL) has been proposed as a privacy-preserving solution for machine learning. However, recent works have reported that FL can leak private client data through membership inference attacks. In this paper, we show that the effectiveness of these attacks on the clients negatively correlates with the size of the client's datasets and model complexity. Based on this finding, we study the capabilities of model-agnostic Federated Learning to preserve privacy, as it enables the use of models of varying complexity in the clients. To systematically study this topic, we first propose a taxonomy of model-agnostic FL methods according to the strategies adopted by the clients to select the sub-models from the server's model. This taxonomy provides a framework for existing model-agnostic FL approaches and leads to the proposal of new FL methods to fill the gaps in the taxonomy. Next, we analyze the privacy-performance trade-off of all the model-agnostic FL architectures as per the proposed taxonomy when subjected to 3 different membership inference attacks on the CIFAR-10 and CIFAR-100 vision datasets. In our experiments, we find that randomness in the strategy used to select the server's sub-model to train the clients' models can control the clients' privacy while keeping competitive performance on the server's side.

High-order Graph Neural Networks (HO-GNNs) have been developed to infer consistent latent spaces in the heterophilic regime, where the label distribution is not correlated with the graph structure. However, most of the existing HO-GNNs are hop-based, i.e., they rely on the powers of the transition matrix. As a result, these architectures are not fully reactive to the classification loss and the achieved structural filters have static supports. In other words, neither the filters' supports nor their coefficients can be learned with these networks. They are confined, instead, to learn combinations of filters. To address the above concerns, we propose Diffusion-jump GNNs a method relying on asymptotic diffusion distances that operates on jumps. A diffusion-pump generates pairwise distances whose projections determine both the support and coefficients of each structural filter. These filters are called jumps because they explore a wide range of scales in order to find bonds between scattered nodes with the same label. Actually, the full process is controlled by the classification loss. Both the jumps and the diffusion distances react to classification errors (i.e. they are learnable). Homophiliation, i.e., the process of learning piecewise smooth latent spaces in the heterophilic regime, is formulated as a Dirichlet problem: the known labels determine the border nodes and the diffusion-pump ensures a minimal deviation of the semi-supervised grouping from a canonical unsupervised grouping. This triggers the update of both the diffusion distances and, consequently, the jumps in order to minimize the classification error. The Dirichlet formulation has several advantages. It leads to the definition of structural heterophily, a novel measure beyond edge heterophily. It also allows us to investigate links with (learnable) diffusion distances, absorbing random walks and stochastic diffusion.

Human perception, memory and decision-making are impacted by tens of cognitive biases and heuristics that influence our actions and decisions. Despite the pervasiveness of such biases, they are generally not leveraged by today's Artificial Intelligence (AI) systems that model human behavior and interact with humans. In this theoretical paper, we claim that the future of human-machine collaboration will entail the development of AI systems that model, understand and possibly replicate human cognitive biases. We propose the need for a research agenda on the interplay between human cognitive biases and Artificial Intelligence. We categorize existing cognitive biases from the perspective of AI systems, identify three broad areas of interest and outline research directions for the design of AI systems that have a better understanding of our own biases., Comment: 14 pages, 1 figure; Accepted for presentation at the AAAI Fall Symposium 2022 on Thinking Fast and Slow and Other Cognitive Theories in AI. Corrected typos; v3: Updated figure 1, added table 4

Federated learning (FL) has been proposed as a privacy-preserving approach in distributed machine learning. A federated learning architecture consists of a central server and a number of clients that have access to private, potentially sensitive data. Clients are able to keep their data in their local machines and only share their locally trained model's parameters with a central server that manages the collaborative learning process. FL has delivered promising results in real-life scenarios, such as healthcare, energy, and finance. However, when the number of participating clients is large, the overhead of managing the clients slows down the learning. Thus, client selection has been introduced as a strategy to limit the number of communicating parties at every step of the process. Since the early na\"{i}ve random selection of clients, several client selection methods have been proposed in the literature. Unfortunately, given that this is an emergent field, there is a lack of a taxonomy of client selection methods, making it hard to compare approaches. In this paper, we propose a taxonomy of client selection in Federated Learning that enables us to shed light on current progress in the field and identify potential areas of future research in this promising area of machine learning., Comment: 17 pages, 3 figures, 1 appendix, accepted to TMLR

We interpret the lack of large-angle temperature correlations and the apparent even-odd parity imbalance, observed in the cosmic microwave background by COBE, WMAP and Planck satellite missions, as a possible stringy signal ultimately stemming from a composite inflaton field (e.g. a fermionic condensate). Based on causality arguments and a Fourier analysis of the angular two-point correlation function, two infrared cutoffs $k_{\rm min}^{\rm even,odd}$ are introduced in the CMB power spectrum associated, respectively, with periodic and antiperiodic boundary conditions of the fermionic constituents (echoing the Neveu-Schwarz-Ramond model in superstring theory), without resorting to any particular model., Comment: 11 pages, 7 figures

This paper proposes an interactive system for mobile devices controlled by hand gestures aimed at helping people with visual impairments. This system allows the user to interact with the device by making simple static and dynamic hand gestures. Each gesture triggers a different action in the system, such as object recognition, scene description or image scaling (e.g., pointing a finger at an object will show a description of it). The system is based on a multi-head neural network architecture, which initially detects and classifies the gestures, and subsequently, depending on the gesture detected, performs a second stage that carries out the corresponding action. This multi-head architecture optimizes the resources required to perform different tasks simultaneously, and takes advantage of the information obtained from an initial backbone to perform different processes in a second stage. To train and evaluate the system, a dataset with about 40k images was manually compiled and labeled including different types of hand gestures, backgrounds (indoors and outdoors), lighting conditions, etc. This dataset contains synthetic gestures (whose objective is to pre-train the system in order to improve the results) and real images captured using different mobile phones. The results obtained and the comparison made with the state of the art show competitive results as regards the different actions performed by the system, such as the accuracy of classification and localization of gestures, or the generation of descriptions for objects and scenes.

The analysis of azimuthal correlations in multiparticle production can be useful to uncover the existence of new physics beyond the Standard Model, e.g., Hidden Valley, in $e^+e^-$ annihilation at high energies. In this paper, based on previous theoretical studies and using PYTHIA8 event generator, it is found that both azimuthal and rapidity long-range correlations are enhanced due to the presence of a new stage of matter on top of the QCD partonic cascade. Ridge structures, similar to those observed at the LHC, show up providing a possible signature of new physics at future $e^+e^-$ colliders., Comment: 8 pages, 7 figures

In this paper, analogies between multiparticle production in high-energy collisions and the time evolution of the early universe are discussed. A common explanation is put forward under the assumption of an unconventional early state: a rapidly expanding universe before recombination (last scattering surface), followed by the CMB, later evolving up to present days, versus the formation of hidden/dark states in hadronic collisions followed by a conventional QCD parton shower yielding final-state particles. In particular, long-range angular correlations are considered pointing out deep connections between the two physical cases potentially useful for the discovery of new physics., Comment: 9 pages, 5 figures

In this paper, we consider the possibility that a new stage of matter, stemming from hidden/dark sectors beyond the Standard Model, to be formed in $pp$ collisions at the LHC, can significantly modify the correlations among final-state particles. In particular, two-particle azimuthal correlations are studied by means of a Fourier series sensitive to the near-side ridge effect while assuming that hidden/dark particles decay on top of the conventional parton shower. Then, new (fractional) harmonic terms should be included in the Fourier analysis of the azimuthal anisotropies, encoding the hypothetical new physics contribution enabling its detection in a complementary way to other signatures., Comment: 7 pages, 3 figures

A numerical study on the distributions of primes in short intervals of length $h$ over the natural numbers $N$ is presented. Based on Cram\'er's model in Number Theory, we obtain a heuristic expression applicable when $h \gg \log{N}$ but $h \ll N$, providing support to the Montgomery and Soundararajan conjecture on the variance of the prime distribution at this scale., Comment: 8 pages, 6 figures

New phenomena involving pseudorapidity and azimuthal correlations among final state particles in $pp$ collisions at the LHC can hint at the existence of hidden sectors beyond the Standard Model. In this paper we rely on a correlated-cluster picture of multiparticle production, which was shown to account for the ridge effect, to assess the effect of a hidden sector on three-particle correlations concluding that there is a potential signature of new physics that can be directly tested by experiments using well-known techniques., Comment: 6 pages, 4 figures

Pseudorapidity and azimuthal three-particle correlations are studied based on a correlated-cluster model of multiparticle production. The model provides a common framework for correlations in proton-proton and heavy-ion collisions allowing easy comparison with the measurements. It is shown that azimuthal cluster correlations are definitely required in order to understand three-particle correlations in the near-side ridge effect. This is similar to the explanation of the ridge phenomenon found in our previous analysis of two-particle correlations and generalizes the model to higher-order correlations., Comment: 16 pages, 7 figures. arXiv admin note: text overlap with arXiv:1610.06408

A study of the near-side ridge phenomenon in hadron-hadron collisions based on a cluster picture of multiparticle production is presented. The near-side ridge effect is shown to have a natural explanation in this context provided that clusters are produced in a correlated manner in the collision transverse plane., Comment: 1 figure

We study the impact of a hidden sector beyond the Standard Model, e.g. a Hidden Valley model, on factorial moments and cumulants of multiplicity distributions in multiparticle production with a special emphasis on the prospects for LHC results., Comment: 13 pages, 7 figures

Most signatures of new physics in colliders have been studied so far on the transverse plane with respect to the beam direction. In this work however we study the impact of a hidden sector beyond the Standard Model (SM) on inclusive (pseudo)rapidity correlations and moments of the multiplicity distributions, with special emphasis in the LHC results., Comment: Presentation given at ICHEP 2014 Valencia

In this work we estimate the helicity suppressed decay rates of $\eta_b$ resonances into baryon pairs due to instanton-induced effects by rescaling the corresponding partial widths of the experimentally measured branching ratios for the $\eta_c(1S) \to p\bar{p}$ and $\eta_c(1S) \to \Lambda\bar(\Lambda)$ decay modes. Thus we point out that both $\eta_b(1S) \to p\bar{p}$ and $\eta_b(1S) \to \Lambda\bar(\Lambda)$ channels could be detected at a Super B factory and LHC experiments. Furthermore, we examine related instanton-induced effects on WIMP scattering off nuclei concluding, albeit with large uncertainties, that they might enhance the spin-dependent cross section for a light pseudoscalar Higgs boson mediator, thereby inducing a dependence on the momentum transfer to the recoling nucleus., Comment: Presented at Xth Quark Confinement and the Hadron Spectrum Conference, Garching, Munich, October 2012. 8 pages with 4 figures

Motivated by the Goldbach conjecture in Number Theory and the abelian bosonization mechanism on a cylindrical two-dimensional spacetime we study the reconstruction of a real scalar field as a product of two real fermion (so-called \textit{prime}) fields whose Fourier expansion exclusively contains prime modes. We undertake the canonical quantization of such prime fields and construct the corresponding Fock space by introducing creation operators $b_{p}^{\dag}$ --labeled by prime numbers $p$-- acting on the vacuum. The analysis of our model, based on the standard rules of quantum field theory and the assumption of the Riemann hypothesis, allow us to prove that the theory is not renormalizable. We also comment on the potential consequences of this result concerning the validity or breakdown of the Goldbach conjecture for large integer numbers., Comment: 20 pages in A4 format, 2 figures

We explore a scenario in the Next-to-Minimal-Supersymmetric-Standard-Model (NMSSM) with both a light O(10) GeV neutralino and a CP-odd Higgs boson with significant coupling to down-type fermions, evading all current B physics, LEP and WMAP bounds. Motivated by a possible slight lepton universality breaking hinted in Upsilon decays, we consider the effect of the mixing of eta(b) resonances with the pseudoscalar Higgs on the spin-dependent scattering neutralino cross section off nucleons. We conclude that this mechanism could be relevant provided that non-perturbative effects enhance the effective eta(b)-nucleon coupling, taking over velocity/q^2 suppression factors, perhaps giving a new insight into the current controversial situation concerning direct search experiments of dark matter., Comment: We have incorporated a discussion on instanton-induced effects as a possible non-perturbative contribution to WIMP scattering off nuclei. 6 pages, 2 figures

In this paper, we present a more detailed version of our previous work for three-particle correlations in quark and gluon jets [1]. We give theoretical results for this observable in the double logarithmic approximation and the modified leading logarithmic approximation. In both resummation schemes, we use the formalism of the generating functional and solve the evolution equations analytically from the steepest descent evaluation of the one-particle distribution. In addition, in this paper we include predictions beyond the limiting spectrum approximation and study this observable near the hump of the single inclusive distribution. We thus provide a further test of the local parton hadron duality (LPHD) and make predictions for the LHC. The computation of higher rank correlators is presented in the double logarithmic approximation and shown to be rather cumbersome., Comment: 34 pages and 14 figures

Three-particle correlations in quark and gluon jets are computed for the first time in perturbative QCD. We give results in the double logarithmic approximation and the modified leading logarithmic approximation. In both resummation schemes, we use the formalism of the generating functional and solve the evolution equations analytically from the steepest descent evaluation of the one-particle distribution. We thus provide a further test of the local parton hadron duality and make predictions for the LHC., Comment: 9 pages and 5 figures. Version published by Physical Review D with reference: Phys. Rev. D 84, 034015 (2011). Two more figures and one section added

The after-dinner talk has by now become a tradition of this Conference series on Quark Confinement and the Hadron Spectrum. On this occasion, I have tried to combine a free-style and (hopefully) amusing presentation with deep questions of physics especially connected with the dynamics of strong interaction. To this end some masterpieces of classical music (by Beethoven, Mozart, Dvorak, Stravinsky ...) and pop music (by Bob Dylan, Eric Clapton) were employed to illustrate certain aspects of physics. By no means was this presentation (nor this paper) intended as a comprehensive review of the different topics examined during the Conference, but rather a call for further thinking on the sinergy of different branches of physics and the excitement of foreseen discoveries in a not too distant future., Comment: 6 pages, 8 figures, After-dinner talk given at the IX Conference on Quark Confinement and the Hadron Structure, Madrid, August 30th to September 3rd 2010

After inserting the heavy quark mass dependence into QCD partonic evolution equations, we determine the mean charged hadron multiplicity and second multiplicity correlators of jets produced in high energy collisions. We thereby extend the so-called dead cone effect to the phenomenology of multiparticle production in QCD jets and find that the average multiplicity of heavy-quark initiated jets decreases significantly as compared to the massless case, even taking into account the weak decay products of the leading primary quark. We emphasize the relevance of our study as a complementary check of $b$-tagging techniques at hadron colliders like the Tevatron and the LHC., Comment: Version revised, accepted for publication in JHEP, 21 pages and 7 figures

In two recent papers (arXiv:1002.1971 [hep-ph] and arXiv:0911.2460 [hep-ph]) Dermisek and Gunion provide new constraints on a light CP-odd Higgs boson in the framework of the NMSSM (or similar models) based on experimental data from LEP, CLEO, BaBar and CDF experiments. In this brief comment we argue that special care is still needed inside a narrow mass window where mixing with $\eta_b$ resonances below $B\bar{B}$ can occur. We also stress that observables testing lepton universality and a possible distorsion of the bottomonium mass spectrum can provide an alternative analysis at (Super) B-factories in the search of such an elusive light pseudoscalar Higgs-like object., Comment: 2 pages, 3 eps figures, references and comments added

The study of multiparticle dynamics in hadron-hadron collisions at Tevatron and LHC could provide useful information on new physics in addition to the expected signatures on the transverse plane. We suggest that an analysis of inclusive correlations between emitted particles in $pp$ inelastic collisions, and factorial moments of multiplicity distributions, may be helpful in uncovering (un)particles from Hidden Sectors, using underlying events tagged by hard products like high $p_T$ leptons and photons, and applying stringent selection criteria like event shape variables, etc, Comment: 4 pages, 2 figs, Talk given at SUSY09, Northeastern University, Boston June 2009

The mass of the eta_b(1S), measured recently by BABAR, is significantly lower than expected from QCD predictions for the Upsilon(1S) - eta_b(1S) hyperfine splitting. We suggest that the observed eta_b(1S) mass is shifted downwards due to a mixing with a CP-odd Higgs scalar A with a mass m_A in the range 9.4 - 10.5 GeV compatible with LEP, CLEO and BABAR constraints. We determine the resulting predictions for the spectrum of the eta_b(nS) - A system and the branching ratios into tau^+ tau^- as functions of m_A., Comment: 4 pages, 4 figures, submitted to PRL, misprint corrected

Most signatures of new physics have been studied on the transverse plane with respect to the beam direction at the LHC where background is much reduced. In this paper we propose the analysis of inclusive longitudinal (pseudo)rapidity correlations among final-state (charged) particles in order to search for (un)particles belonging to a Hidden Sector beyond the Standard Model, using a selected sample of p-p minimum bias events (applying appropriate off-line cuts on events based on, e.g., minijets, high-multiplicity, event shape variables, high-pT leptons and photons, etc) collected at the early running of the LHC. To this aim, we examine inclusive and semi-inclusive two-particle correlation functions, forward-backward correlations, and factorial moments of the multiplicity distribution, without resorting to any particular model but under very general (though simplifying) assumptions. Finally, motivated by some analysis techniques employed in the search for Quark-Gluon-Plasma in heavy-ion collisions, we investigate the impact of such intermediate (un)particle stuff on the (multi)fractality of parton cascades in p-p collisions, by means of a Levy stable law description and a Ginzburg-Landau model of phase transitions. Results from our preliminary study seem encouraging for possible dedicated analyses at the LHC and Tevatron experiments., Comment: 36 pages, 6 figures. New references and a new section on Glasma physics have been added

We study possible effects of a light CP-odd Higgs boson on radiative Upsilon decays in the Next-to-Minimal Supersymmetric Standard Model. Recent constraints from CLEO on radiative Upsilon(1S) decays are translated into constraints on the parameter space of CP-odd Higgs boson masses and couplings, and compared to constraints from B physics and the muon anomalous magnetic moment. Possible Higgs - eta_b(nS) mixing effects are discussed, notably in the light of the recent measurement of the eta_b(1S) mass by Babar: The somewhat large Upsilon(1S) - eta_b(1S) hyperfine splitting could easily be explained by the presence of a CP-odd Higgs boson with a mass in the range 9.4 - 10.5 GeV. Then, tests of lepton universality in inclusive radiative Upsilon decays can provide a visible signal in forthcoming experimental data., Comment: 20 pages, 11 Figs, references corrected

We examine the electronic width ratios of Upsilon resonances below the BBbar threshold by means of an effective (Cornell-type) QCD potential incorporating 1/m_b corrections obtained from a prior fit to the bottomonium spectrum. From our analysis we conclude that the Upsilon(2S) and Upsilon(3S) states should belong to the strong-coupling (nonperturbative) regime while the Upsilon(1S) state should belong to the weak-coupling (perturbative) regime, in agreement with a previous study based on radiative decays., Comment: 8 pages

This paper presents the basics of the QQ-onia package, a software based upon the Numerov method which can be used to solve the Schrodinger radial equation using a suitable potential V(r) for the heavy quarkonium system. This package also allows the analysis of relevant properties of those resonances such as the wave functions at the origin, their corresponding derivatives for l \neq 0 states, typical heavy quark velocities, and mean square radii. Besides, it includes a tool to analize the spin-dependent contributions to the heavy quarkonia spectrum, providing the hyperfine splittings, as well as the nPJ energy levels. Finally, a simple software implemented in QQ-onia to compute E1 transition rates is presented., Comment: 30 pages, 3 figures, a new section on spin-dependent terms in the potential has been added

A light non-standard Higgs boson decaying mainly into $\tau^+\tau^-$ has not been yet ruled out by LEP searches in several scenarios. We verify that, in the context of the Next-to-Minimal-Supersymmetric Standard Model, a low-mass CP-odd (mostly but not completely singlet-like) Higgs boson can couple strongly enough to down-type fermions to be detected in radiative $\Upsilon$ decays into tauonic pairs at a high-luminosity B factory. Possible spectroscopic effects of a mixing with $\eta_b$ resonances are also analyzed, Comment: Latex, 4 pages, 4 figures. Submitted for the SUSY07 proceedings

Several scenarios beyond the minimal extension of the Standard Model still allow light non-standard Higgs bosons evading LEP bounds. We examine the mixing between a light CP-odd Higgs boson and $\eta_b$ states and its implications on a slight (but observable) lepton universality breaking in Upsilon decays which could be measured at the percent level at a (Super) B factory., Comment: 9 pages, 7 figures

We present a proposal for detecting new physics at a B-factory running at the $\Upsilon(3S)$ resonance by testing lepton universality to the few percent level in the leptonic decays of the $\Upsilon(1S)$ and $\Upsilon(2S)$ resonances tagged by the dipion in the chain decay: $\Upsilon(3S) \to pi^+\pi^-\Upsilon(1S,2S)$; $\Upsilon(1S,2S) \to \ell^+\ell^-$, $\ell=e,\mu,\tau$., Comment: 10 pages, 2 figures, Contributed paper to BNM2006 Workshop, September 13-14, 2006, KEK

A non-standard CP-odd Higgs boson could induce a slight (but observable) lepton universality breaking in Upsilon leptonic decays. Moreover, mixing between such a pseudoscalar Higgs boson and $\eta_b$ states might shift their mass levels, thereby modifying the values of the $M_{\Upsilon(nS)}-M_{\eta_b(nS)}$ hyperfine spplitings predicted in the standard model. Besides, $\eta_b$ resonances could be broader than expected with potentially negative consequences for discovery in both $e^+e^-$ and hadron colliders. A scenario with a CP violating Higgs sector is also considered. Finally, further strategies to search for a light Higgs particle in bottomonium decays are outlined., Comment: LaTeX, 10 pages, 1 EPS figure