Search

Current interactive segmentation approaches, inspired by the success of META's Segment Anything model, have achieved notable advancements, however, they come with substantial limitations that hinder their practical application in real clinical scenarios. These include unrealistic human interaction requirements, such as slice-by-slice operations for 2D models on 3D data, a lack of iterative refinement, and insufficient evaluation experiments. These shortcomings prevent accurate assessment of model performance and lead to inconsistent outcomes across studies. IntRaBench overcomes these challenges by offering a comprehensive and reproducible framework for evaluating interactive segmentation methods in realistic, clinically relevant scenarios. It includes diverse datasets, target structures, and segmentation models, and provides a flexible codebase that allows seamless integration of new models and prompting strategies. Additionally, we introduce advanced techniques to minimize clinician interaction, ensuring fair comparisons between 2D and 3D models. By open-sourcing IntRaBench, we invite the research community to integrate their models and prompting techniques, ensuring continuous and transparent evaluation of interactive segmentation models in 3D medical imaging., Comment: Undergoing Peer-Review

Social contact surveys are an important tool to assess infection risks within populations, and the effect of non-pharmaceutical interventions on social behaviour during disease outbreaks, epidemics, and pandemics. Numerous longitudinal social contact surveys were conducted during the COVID-19 era, however data analysis is plagued by reporting fatigue, a phenomenon whereby the average number of social contacts reported declines with the number of repeat participations and as participants' engagement decreases over time. Using data from the German COVIMOD Study between April 2020 to December 2021, we demonstrate that reporting fatigue varied considerably by sociodemographic factors and was consistently strongest among parents reporting children contacts (parental proxy reporting), students, middle-aged individuals, those in full-time employment and those self-employed. We find further that, when using data from first-time participants as gold standard, statistical models incorporating a simple logistic function to control for reporting fatigue were associated with substantially improved estimation accuracy relative to models with no reporting fatigue adjustments, and that no cap on the number of repeat participations was required. These results indicate that existing longitudinal contact survey data can be meaningfully interpreted under an easy-to-implement statistical approach adressing reporting fatigue confounding, and that longitudinal designs including repeat participants are a viable option for future social contact survey designs.

How can we test AI performance? This question seems trivial, but it isn't. Standard benchmarks often have problems such as in-distribution and small-size test sets, oversimplified metrics, unfair comparisons, and short-term outcome pressure. As a consequence, good performance on standard benchmarks does not guarantee success in real-world scenarios. To address these problems, we present Touchstone, a large-scale collaborative segmentation benchmark of 9 types of abdominal organs. This benchmark is based on 5,195 training CT scans from 76 hospitals around the world and 5,903 testing CT scans from 11 additional hospitals. This diverse test set enhances the statistical significance of benchmark results and rigorously evaluates AI algorithms across various out-of-distribution scenarios. We invited 14 inventors of 19 AI algorithms to train their algorithms, while our team, as a third party, independently evaluated these algorithms on three test sets. In addition, we also evaluated pre-existing AI frameworks--which, differing from algorithms, are more flexible and can support different algorithms--including MONAI from NVIDIA, nnU-Net from DKFZ, and numerous other open-source frameworks. We are committed to expanding this benchmark to encourage more innovation of AI algorithms for the medical domain., Comment: Accepted to NeurIPS-2024

Self-Supervised Learning (SSL) presents an exciting opportunity to unlock the potential of vast, untapped clinical datasets, for various downstream applications that suffer from the scarcity of labeled data. While SSL has revolutionized fields like natural language processing and computer vision, their adoption in 3D medical image computing has been limited by three key pitfalls: Small pre-training dataset sizes, architectures inadequate for 3D medical image analysis, and insufficient evaluation practices. We address these issues by i) leveraging a large-scale dataset of 44k 3D brain MRI volumes and ii) using a Residual Encoder U-Net architecture within the state-of-the-art nnU-Net framework. iii) A robust development framework, incorporating 5 development and 8 testing brain MRI segmentation datasets, allowed performance-driven design decisions to optimize the simple concept of Masked Auto Encoders (MAEs) for 3D CNNs. The resulting model not only surpasses previous SSL methods but also outperforms the strong nnU-Net baseline by an average of approximately 3 Dice points. Furthermore, our model demonstrates exceptional stability, achieving the highest average rank of 2 out of 7 methods, compared to the second-best method's mean rank of 3., Comment: Arxiv Preprint. Currently under Review

Ethanolamine (NH$_2$CH$_2$CH$_2$OH, EA) has been identified in the gas phase of the ISM within molecular clouds. Although EA has not been directly observed in the molecular ice phase, a solid state formation mechanism has been proposed. However, the current literature lacks an estimation of the infrared band strengths of EA ices. We conducted an experimental investigation of solid EA ice at low temperatures to ascertain its infrared band strengths, phase transition temperature, and multilayer binding energy. The commonly used laser interferometry method was not applied. Infrared band strengths were determined using three distinct methods. The obtained lab spectrum of EA was compared with the publicly available MIRI MRS James Webb Space Telescope observations toward a low mass protostar. The phase transition temperature for EA ice falls within the range of 175 to 185 K. Among the discussed methods, the simple pressure gauge method provides a reasonable estimate of band strength. We derive a band strength value of about $1\times10^{-17}$ cm molecule$^{-1}$ for the NH$_2$ bending mode in the EA molecules. Additionally, temperature-programmed desorption analysis yielded a multilayer desorption energy of 0.61$\pm$0.01 eV. By comparing the laboratory data documented in this study with the JWST spectrum of the low mass protostar IRAS 2A, an upper-limit for the EA ice abundances was derived., Comment: 8 pages, 5 figures

Contemporary machine learning methods will try to approach the Bayes error, as it is the lowest possible error any model can achieve. This paper postulates that any decision is composed of not one but two Bayesian decisions and that decision-making is, therefore, a double-Bayesian process. The paper shows how this duality implies intrinsic uncertainty in decisions and how it incorporates explainability. The proposed approach understands that Bayesian learning is tantamount to finding a base for a logarithmic function measuring uncertainty, with solutions being fixed points. Furthermore, following this approach, the golden ratio describes possible solutions satisfying Bayes' theorem. The double-Bayesian framework suggests using a learning rate and momentum weight with values similar to those used in the literature to train neural networks with stochastic gradient descent., Comment: 14 pages, 5 figures, draft

Type II supernovae (SNeII) mark the endpoint in the lives of hydrogen-rich massive stars. Their large explosion energies and luminosities allow us to measure distances, metallicities, and star formation rates into the distant Universe. To fully exploit their use in answering different astrophysical problems, high-quality low-redshift data sets are required. Such samples are vital to understand the physics of SNeII, but also to serve as calibrators for distinct - and often lower-quality - samples. We present uBgVri optical and YJH near-infrared (NIR) photometry for 94 low-redshift SNeII observed by the Carnegie Supernova Project (CSP). A total of 9817 optical and 1872 NIR photometric data points are released, leading to a sample of high-quality SNII light curves during the first ~150 days post explosion on a well-calibrated photometric system. The sample is presented and its properties are analysed and discussed through comparison to literature events. We also focus on individual SNeII as examples of classically defined subtypes and outlier objects. Making a cut in the plateau decline rate of our sample (s2), a new subsample of fast-declining SNeII is presented. The sample has a median redshift of 0.015, with the nearest event at 0.001 and the most distant at 0.07. At optical wavelengths (V), the sample has a median cadence of 4.7 days over the course of a median coverage of 80 days. In the NIR (J), the median cadence is 7.2 days over the course of 59 days. The fast-declining subsample is more luminous than the full sample and shows shorter plateau phases. Of the non-standard SNeII highlighted, SN2009A particularly stands out with a steeply declining then rising light curve, together with what appears to be two superimposed P-Cygni profiles of H-alpha in its spectra. We outline the significant utility of these data, and finally provide an outlook of future SNII science., Comment: Accepted for publication in A&A. Photometric data will be uploaded to the CDS and the CSP website, and can also be requested from the first author

Inverse design (ID) is a computational method that systematically explores a design space to find optimal device geometries based on specific performance criteria. In silicon photonics, ID often leads to devices with design features that degrade significantly due to the fabrication process, limiting the applicability of these devices in scalable silicon photonic fabrication. We demonstrate a solution to this performance degradation through fabrication-aware inverse design (FAID), integrating lithography models for deep-ultraviolet (DUV) lithography and electron beam lithography (EBL) into the shape optimization approach of ID. A Y-branch and an SWG-to-strip converter were generated and fabricated with this new approach. Simulated and measured results verify that the FAID yields devices with up to 0.6 dB lower insertion loss per device. The modified workflow enables designers to use ID to generate devices that adjust for process bias predicted by lithography models., Comment: 4 pages

We measure the response of open-cell polyurethane foams filled with a dense suspension of fumed silica particles in polyethylene glycol at compression speeds spanning several orders of magnitude. The gradual compressive stress increase of the composite material indicates the existence of shear rate gradients in the interstitial suspension caused by wide distributions in pore sizes in the disordered foam network. The energy dissipated during compression scales with an effective internal shear rate, allowing for the collapse of three data sets for different pore-size foams. When scaled by this effective shear rate, the most pronounced energy increase coincides with the effective shear rate corresponding to the onset of shear thickening in our bulk suspension. Optical measurements of the radial deformation of the foam network and of the suspension flow under compression provide additional insight into the interaction between shear thickening fluid and foam. This optical data, combined with a simple model of a spring submerged in viscous flow, illustrates the dynamic interaction of viscous drag with foam elasticity as a function of compression rate, and identifies the foam pore size distribution as a critically important model parameter. Taken together, the stress measurements, dissipated energy, and relative motion of the fluid and the foam can be rationalized by knowing the pore size distribution and the average pore size of the foam.

The various limitations of Generative AI, such as hallucinations and model failures, have made it crucial to understand the role of different modalities in Visual Language Model (VLM) predictions. Our work investigates how the integration of information from image and text modalities influences the performance and behavior of VLMs in visual question answering (VQA) and reasoning tasks. We measure this effect through answer accuracy, reasoning quality, model uncertainty, and modality relevance. We study the interplay between text and image modalities in different configurations where visual content is essential for solving the VQA task. Our contributions include (1) the Semantic Interventions (SI)-VQA dataset, (2) a benchmark study of various VLM architectures under different modality configurations, and (3) the Interactive Semantic Interventions (ISI) tool. The SI-VQA dataset serves as the foundation for the benchmark, while the ISI tool provides an interface to test and apply semantic interventions in image and text inputs, enabling more fine-grained analysis. Our results show that complementary information between modalities improves answer and reasoning quality, while contradictory information harms model performance and confidence. Image text annotations have minimal impact on accuracy and uncertainty, slightly increasing image relevance. Attention analysis confirms the dominant role of image inputs over text in VQA tasks. In this study, we evaluate state-of-the-art VLMs that allow us to extract attention coefficients for each modality. A key finding is PaliGemma's harmful overconfidence, which poses a higher risk of silent failures compared to the LLaVA models. This work sets the foundation for rigorous analysis of modality integration, supported by datasets specifically designed for this purpose.

Medical imaging is spearheading the AI transformation of healthcare. Performance reporting is key to determine which methods should be translated into clinical practice. Frequently, broad conclusions are simply derived from mean performance values. In this paper, we argue that this common practice is often a misleading simplification as it ignores performance variability. Our contribution is threefold. (1) Analyzing all MICCAI segmentation papers (n = 221) published in 2023, we first observe that more than 50% of papers do not assess performance variability at all. Moreover, only one (0.5%) paper reported confidence intervals (CIs) for model performance. (2) To address the reporting bottleneck, we show that the unreported standard deviation (SD) in segmentation papers can be approximated by a second-order polynomial function of the mean Dice similarity coefficient (DSC). Based on external validation data from 56 previous MICCAI challenges, we demonstrate that this approximation can accurately reconstruct the CI of a method using information provided in publications. (3) Finally, we reconstructed 95% CIs around the mean DSC of MICCAI 2023 segmentation papers. The median CI width was 0.03 which is three times larger than the median performance gap between the first and second ranked method. For more than 60% of papers, the mean performance of the second-ranked method was within the CI of the first-ranked method. We conclude that current publications typically do not provide sufficient evidence to support which models could potentially be translated into clinical practice., Comment: Paper accepted at MICCAI 2024 conference

Explainable AI (XAI) is a rapidly growing domain with a myriad of proposed methods as well as metrics aiming to evaluate their efficacy. However, current studies are often of limited scope, examining only a handful of XAI methods and ignoring underlying design parameters for performance, such as the model architecture or the nature of input data. Moreover, they often rely on one or a few metrics and neglect thorough validation, increasing the risk of selection bias and ignoring discrepancies among metrics. These shortcomings leave practitioners confused about which method to choose for their problem. In response, we introduce LATEC, a large-scale benchmark that critically evaluates 17 prominent XAI methods using 20 distinct metrics. We systematically incorporate vital design parameters like varied architectures and diverse input modalities, resulting in 7,560 examined combinations. Through LATEC, we showcase the high risk of conflicting metrics leading to unreliable rankings and consequently propose a more robust evaluation scheme. Further, we comprehensively evaluate various XAI methods to assist practitioners in selecting appropriate methods aligning with their needs. Curiously, the emerging top-performing method, Expected Gradients, is not examined in any relevant related study. LATEC reinforces its role in future XAI research by publicly releasing all 326k saliency maps and 378k metric scores as a (meta-)evaluation dataset. The benchmark is hosted at: https://github.com/IML-DKFZ/latec., Comment: Accepted at NeurIPS 2024

Luminescent coupling is a characteristic of multi-junction solar cells which has often been neglected in models of their performance. The effect describes the absorption of light emitted from a higher band gap semiconductor by a lower band gap semiconductor. In this way, light which might have been lost can be utilized for current generation. We present a framework for modeling this effect in both planar layer stacks and devices with periodic nanostructuring. As a case study, we evaluate how luminescent coupling is affected by the inclusion of nanostructuring in a perovskite silicon tandem solar cell. We find that nanostructuring, while reducing the reflection loss for tandem solar cells also reduces the luminescent coupling, allowing more light to be emitted to the surroundings, when compared to planar devices. This highlights the need to include modeling of this effect into optimization schemes in order to find the trade-off between these two effects. The published version of this work is available at https://doi.org/10.1117/12.3023941., Comment: 7 pages, 3 figures

In this note, we generalise the concept of topo-isomorphic extensions and define finite topomorphic extensions as topological dynamical systems whose factor map to the maximal equicontinuous factor is measure-theoretically at most $m$-to-one for some $m\in\mathbb{N}$. We further define multivariate versions of mean equicontinuity, complementing the notion of multivariate mean sensitivity introduced by Li, Ye and Yu, and then show that any $m$-to-one topomorphic extension is mean $(m+1)$-equicontinuous. This falls in line with the well-known result, due to Downarowicz and Glasner, that strictly ergodic systems are isomorphic extensions if and only if they are mean equicontinuous. While in the multivariate case we can only conjecture that the converse direction also holds, the result provides an indication that multivariate equicontinuity properties are strongly related to finite extension structures. For minimal systems, an Auslander-Yorke type dichotomy between multivariate mean equicontinuity and sensitivity is shown as well.

We introduce ExcitationSolve, a fast globally-informed gradient-free optimizer for physically-motivated ans\"atze constructed of excitation operators, a common choice in variational quantum eigensolvers. ExcitationSolve is to be classified as an extension of quantum-aware and hyperparameter-free optimizers such as Rotosolve, from parameterized unitaries with generators $G$ of the form $G^2=I$, e.g., rotations, to the more general class of $G^3=G$ exhibited by the physically-inspired excitation operators such as in the unitary coupled cluster approach. ExcitationSolve is capable of finding the global optimum along each variational parameter using the same quantum resources that gradient-based optimizers require for a single update step. We provide optimization strategies for both fixed- and adaptive variational ans\"atze, as well as a multi-parameter generalization for the simultaneous selection and optimization of multiple excitation operators. Finally, we demonstrate the utility of ExcitationSolve by conducting electronic ground state energy calculations of molecular systems and thereby outperforming state-of-the-art optimizers commonly employed in variational quantum algorithms. Across all tested molecules in their equilibrium geometry, ExcitationSolve remarkably reaches chemical accuracy in a single sweep over the parameters of a fixed ansatz. This sweep requires only the quantum circuit executions of one gradient descent step. In addition, ExcitationSolve achieves adaptive ans\"atze consisting of fewer operators than in the gradient-based adaptive approach, hence decreasing the circuit execution time., Comment: 45 pages, 10 figures

Quantum optimization has emerged as a promising frontier of quantum computing, providing novel numerical approaches to mathematical optimization problems. The main goal of this paper is to facilitate interdisciplinary research between the Operations Research (OR) and Quantum Computing communities by providing an OR scientist's perspective on selected quantum-powered methods for discrete optimization. To this end, we consider three quantum-powered optimization approaches that make use of different types of quantum hardware available on the market. To illustrate these approaches, we solve three classical optimization problems: the Traveling Salesperson Problem, Weighted Maximum Cut, and Maximum Independent Set. With a general OR audience in mind, we attempt to provide an intuition behind each approach along with key references, describe the corresponding high-level workflow, and highlight crucial practical considerations. In particular, we emphasize the importance of problem formulations and device-specific configurations, and their impact on the amount of resources required for computation (where we focus on the number of qubits). These points are illustrated with a series of experiments on three types of quantum computers: a neutral atom machine from QuEra, a quantum annealer from D-Wave, and a gate-based device from IBM., Comment: 43 pages, 18 figures, 7 tables. Technical supplement: https://alex-bochkarev.github.io/qopt-overview . Source code, problem instances, and other raw data: https://github.com/alex-bochkarev/qopt-overview

In this work we determine the metric dimension of $ K_a \times K_b \times K_c$ for all $a,b,c\in \mathbb N$ with $ a \le b \le c $ as follows. For $3a c$, it is $\left \lfloor \frac{2}{3}(b+c-1) \right \rfloor$, and for $3a=b+c$, it is $\left \lfloor \frac{a+b+c}{2} \right \rfloor -1 $. The only open case is $3a>b+c$, where two values are possible, namely $\left \lfloor \frac{a+b+c}{2} \right \rfloor -1 $ and $\left \lfloor \frac{a+b+c}{2} \right \rfloor $. This result extends previous results of C\'acere et al., who computed the metric dimension of $ K_a \times K_b$, and of Drewes and J\"ager, who computed the metric dimension of $ K_a \times K_a \times K_a$. We prove our result by introducing and analyzing a new variant of Static Black-Peg Mastermind, in which each peg has its own permitted set of colors. For all cases, we present strategies which we prove to be both feasible and optimal. Our main result follows, as the number of questions of these strategies is equal to the metric dimension of $K_a \times K_b \times K_c$., Comment: 41 pages

Software-defined networking (SDN) is a centralized, dynamic, and programmable network management technology that enables flexible traffic control and scalability. SDN facilitates network administration through a centralized view of the underlying physical topology; tampering with this topology view can result in catastrophic damage to network management and security. To underscore this issue, we introduce Marionette, a new topology poisoning technique that manipulates OpenFlow link discovery packet forwarding to alter topology information. Our approach exposes an overlooked yet widespread attack vector, distinguishing itself from traditional link fabrication attacks that tamper, spoof, or relay discovery packets at the data plane. Unlike localized attacks observed in existing methods, our technique introduces a globalized topology poisoning attack that leverages control privileges. Marionette implements a reinforcement learning algorithm to compute a poisoned topology target, and injects flow entries to achieve a long-lived stealthy attack. Our evaluation shows that Marionette successfully attacks five open-source controllers and nine OpenFlow-based discovery protocols. Marionette overcomes the state-of-the-art topology poisoning defenses, showcasing a new class of topology poisoning that initiates on the control plane. This security vulnerability was ethically disclosed to OpenDaylight, and CVE-2024-37018 has been assigned., Comment: 17 pages, 12 figures, CCS 2024

Medical image classification in radiology faces significant challenges, particularly in generalizing to unseen pathologies. In contrast, CLIP offers a promising solution by leveraging multimodal learning to improve zero-shot classification performance. However, in the medical domain, lesions can be small and might not be well represented in the embedding space. Therefore, in this paper, we explore the potential of visual prompt engineering to enhance the capabilities of Vision Language Models (VLMs) in radiology. Leveraging BiomedCLIP, trained on extensive biomedical image-text pairs, we investigate the impact of embedding visual markers directly within radiological images to guide the model's attention to critical regions. Our evaluation on the JSRT dataset, focusing on lung nodule malignancy classification, demonstrates that incorporating visual prompts $\unicode{x2013}$ such as arrows, circles, and contours $\unicode{x2013}$ significantly improves classification metrics including AUROC, AUPRC, F1 score, and accuracy. Moreover, the study provides attention maps, showcasing enhanced model interpretability and focus on clinically relevant areas. These findings underscore the efficacy of visual prompt engineering as a straightforward yet powerful approach to advance VLM performance in medical image analysis., Comment: Accepted at ECCV 2024 Workshop on Emergent Visual Abilities and Limits of Foundation Models

The Polytope Scheduling Problem (PSP) was introduced by Im, Kulkarni, and Munagala (JACM 2018) as a very general abstraction of resource allocation over time and captures many well-studied problems including classical unrelated machine scheduling, multidimensional scheduling, and broadcast scheduling. In PSP, jobs with different arrival times receive processing rates that are subject to arbitrary packing constraints. An elegant and well-known algorithm for instantaneous rate allocation with good fairness and efficiency properties is the Proportional Fairness algorithm (PF), which was analyzed for PSP by Im et al. We drastically improve the analysis of the PF algorithm for both the general PSP and several of its important special cases subject to the objective of minimizing the sum of weighted completion times. We reduce the upper bound on the competitive ratio from 128 to 27 for general PSP and to 4 for the prominent class of monotone PSP. For certain heterogeneous machine environments we even close the substantial gap to the lower bound of 2 for non-clairvoyant scheduling. Our analysis also gives the first polynomial-time improvements over the nearly 30-year-old bounds on the competitive ratio of the doubling framework by Hall, Shmoys, and Wein (SODA 1996) for clairvoyant online preemptive scheduling on unrelated machines. Somewhat surprisingly, we achieve this improvement by a non-clairvoyant algorithm, thereby demonstrating that non-clairvoyance is not a (significant) hurdle. Our improvements are based on exploiting monotonicity properties of PSP, providing tight dual fitting arguments on structured instances, and showing new additivity properties on the optimal objective value for scheduling on unrelated machines. Finally, we establish new connections of PF to matching markets, and thereby provide new insights on equilibria and their computational complexity.

We present a next-to-leading order QCD calculation for the electroweak production of a pair of same-sign weak gauge bosons in association with three jets in the region dominated by vector boson scattering, and its matching with parton-shower programs according to the POWHEG formalism. Our calculation allows for an accurate description of jet observables crucial for the optimization of experimental analyses of vector boson scattering processes. We find QCD corrections and parton-shower effects moderate in size for inclusive quantities and distributions of the tagging jets. A larger impact of these corrections is observed for subleading jets., Comment: 23 pages, 11 figures

Computed ultrasound tomography in echo mode generates maps of tissue speed of sound (SoS) from the shift of echoes when detected under varying steering angles. It solves a linearized inverse problem that requires regularization to complement the echo shift data with a priori constraints. Spatial gradient regularization has been used to enforce smooth solutions, but SoS estimates were found to be biased depending on tissue layer geometry. Here, we propose to train a linear operator to minimize SoS errors on average over a large number of random tissue models that sample the distribution of geometries and SoS values expected in vivo. In an extensive simulation study on liver imaging, we demonstrate that biases are strongly reduced, with residual biases being the result of a partial non-linearity in the actual physical problem. This approach can either be applied directly to echo-shift data or to the SoS maps estimated with gradient regularization, where the former shows slightly better performance, but the latter is computationally more efficient. Experimental phantom results confirm the transferability of our results to real ultrasound data., Comment: This work has been submitted to the IEEE for possible publication

Stress-strain relations for random packings of entangling chains under triaxial compression can exhibit strain stiffening and sustain stresses several orders-of-magnitude beyond typical granular materials. X-ray tomography reveals the transition to this strong strain stiffening occurs when chains are long enough to entangle an average of about one chain each, which results in system-filling clusters of entangled chains. The number of entanglements is nearly proportional to the area surrounded by entangling particles with an excluded volume effect. A tendency was found for chain links to stretch when the packing was strained. The slope of the stress-strain relation of the packing can be calculated from a mean-field model consisting of the product of the effective extensional modulus of the chain, packing fraction, probability of stretched links, and the ratio of strain of stretched links to packing strain. The stress-strain model requires as input measurements of the ratio between local particle deformation and global average strain, and the probability of stretching for non-rigid particles. This results in a quadratic prediction for the stress-strain curve, with a curvature that agrees with experiments within the model uncertainties. This model explains that the strength of these packings comes from stretching of the links of chains, but only when the system-filling network of entanglements provides constraints that prevents failure by shear banding, so that particles must be deformed to move further under strain. In this model, the increasing slope of the stress-strain curve is mainly due to the fraction of stretched links increasing with strain. This model for the stress-strain relation is shown to be generalizable to different shapes of entangling particles by applying it to staples., Comment: 21 pages, 24 figures, submitted to Physical Review E

The Eye Movements on Machine-Generated Texts Corpus (EMTeC) is a naturalistic eye-movements-while-reading corpus of 107 native English speakers reading machine-generated texts. The texts are generated by three large language models using five different decoding strategies, and they fall into six different text type categories. EMTeC entails the eye movement data at all stages of pre-processing, i.e., the raw coordinate data sampled at 2000 Hz, the fixation sequences, and the reading measures. It further provides both the original and a corrected version of the fixation sequences, accounting for vertical calibration drift. Moreover, the corpus includes the language models' internals that underlie the generation of the stimulus texts: the transition scores, the attention scores, and the hidden states. The stimuli are annotated for a range of linguistic features both at text and at word level. We anticipate EMTeC to be utilized for a variety of use cases such as, but not restricted to, the investigation of reading behavior on machine-generated text and the impact of different decoding strategies; reading behavior on different text types; the development of new pre-processing, data filtering, and drift correction algorithms; the cognitive interpretability and enhancement of language models; and the assessment of the predictive power of surprisal and entropy for human reading times. The data at all stages of pre-processing, the model internals, and the code to reproduce the stimulus generation, data pre-processing and analyses can be accessed via https://github.com/DiLi-Lab/EMTeC/.

We study the emergence of dynamical phases of a Bose-Einstein condensate that is optomechanically coupled to a dissipative cavity mode and transversally driven by a laser. We focus on the regime close to the optomechanical resonance, where the atoms' refractive index shifts the cavity into resonance, assuming fast cavity relaxation. We derive an effective model for the atomic motion, where the cavity degrees of freedom are eliminated using perturbation theory in the atom-cavity coupling and benchmark its predictions using numerical simulations based on the full model. Away from the optomechanical resonance, perturbation theory in the lowest order (adiabatic elimination) reliably describes the dynamics and predicts chaotic phases with unstable oscillations. Interestingly, the dynamics close to the optomechanical resonance are qualitatively captured only by including the corrections to next order (non-adiabatic corrections). In this regime we find limit cycle phases that describe stable oscillations of the density with a well defined frequency. We further show that such limit cycle solutions are metastable configurations of the adiabatic model. Our work sheds light on the mechanisms that are required to observe dynamical phases and predict their existence in atom-cavity systems where a substantial timescale separation is present., Comment: 11 pages, 5 figures

The radioactive decay properties of $^{242}$Es were studied with significantly improved statistics compared to available literature data. This isotope was produced in the 3n evaporation channel of the fusion reaction of $^{48}$Ca+$^{197}$Au. A half-life of 16.9(8)~s was deduced from 662 $\alpha$ decays of $^{242}$Es, resulting in an $\alpha$-decay branching of 41(3)\%. Twenty-six fission events with a half-life of 18.2$^{+4.5}_{-3.0}$~s were assigned to originate from the electron-capture delayed fission of $^{242}$Es. The probability for the electron-capture delayed fission was measured to be 0.015(4), which improves and resolves ambiguities in available experimental data. We discuss all known cases for electron-capture delayed fission in Es, Bk, and Am isotopes and compare experimental data with predictions from a recent semi-empirical model. A cross section of 27(3)~nb was measured for the production of $^{242}$Es.

We study the onset of collective spin-self-organization in a thermal ensemble of driven two-level atoms confined in an optical cavity. The atoms spontaneously form a spin-pattern above a critical driving strength that sets a threshold and is determined by the cavity parameters, the initial temperature, and the transition frequency of the atomic spin. Remarkably, we find that inhomogeneous Doppler broadening facilitates the onset of spin-self-organization. In particular, the threshold is non-monotonic when increasing the spin transition frequency and reaches a minimum when the Doppler broadening is of similar magnitude. This feature emerges due to Doppler-induced resonances. Above the threshold, we find cooperative dynamics of spin, spatial, and momentum degrees of freedom leading to density modulations, fast reduction of kinetic energy, and the emergence of non-thermal states. More broadly, our work demonstrates how broadening can facilitate strong light-matter interactions in many-body systems., Comment: 7+7 pages, 3 figures

Atoms coupled to cavities provide an exciting playground for the study of fundamental interactions of atoms mediated through a common channel. Many of the applications of cavity-QED and cold-atom experiments more broadly, suffer from limitations caused by the transient nature of an atomic loading cycle. The development of continuous operation schemes is necessary to push these systems to the next level of performance. Here we present a machine designed to produce a continuous flux of collimated atoms that traverse an optical cavity. The atom-light interaction is enhanced by a fast-decaying cavity optimal for studying phenomena where atomic properties dominate. We demonstrate the transition to a collective strong coupling regime heralded by a normal-mode splitting. We observe a second phase with a binary normal-mode splitting born from an offset in the mean velocity of the atoms. Inverting the atomic ensemble in the collective strong coupling regime, we measure continuous optical gain. This work sets the stage for studying threshold conditions for continuous collective phenomena, such as continuous superradiant lasing.

Matching cancer patients to clinical trials is essential for advancing treatment and patient care. However, the inconsistent format of medical free text documents and complex trial eligibility criteria make this process extremely challenging and time-consuming for physicians. We investigated whether the entire trial matching process - from identifying relevant trials among 105,600 oncology-related clinical trials on clinicaltrials.gov to generating criterion-level eligibility matches - could be automated using Large Language Models (LLMs). Using GPT-4o and a set of 51 synthetic Electronic Health Records (EHRs), we demonstrate that our approach identifies relevant candidate trials in 93.3% of cases and achieves a preliminary accuracy of 88.0% when matching patient-level information at the criterion level against a baseline defined by human experts. Utilizing LLM feedback reveals that 39.3% criteria that were initially considered incorrect are either ambiguous or inaccurately annotated, leading to a total model accuracy of 92.7% after refining our human baseline. In summary, we present an end-to-end pipeline for clinical trial matching using LLMs, demonstrating high precision in screening and matching trials to individual patients, even outperforming the performance of qualified medical doctors. Our fully end-to-end pipeline can operate autonomously or with human supervision and is not restricted to oncology, offering a scalable solution for enhancing patient-trial matching in real-world settings., Comment: 149 pages, including Supplements. 3 Main Figures

Filesystem vulnerabilities persist as a significant threat to Android systems, despite various proposed defenses and testing techniques. The complexity of program behaviors and access control mechanisms in Android systems makes it challenging to effectively identify these vulnerabilities. In this paper, we present PathSentinel, which overcomes the limitations of previous techniques by combining static program analysis and access control policy analysis to detect three types of filesystem vulnerabilities: path traversals, hijacking vulnerabilities, and luring vulnerabilities. By unifying program and access control policy analysis, PathSentinel identifies attack surfaces accurately and prunes many impractical attacks to generate input payloads for vulnerability testing. To streamline vulnerability validation, PathSentinel leverages large language models (LLMs) to generate targeted exploit code based on the identified vulnerabilities and generated input payloads. The LLMs serve as a tool to reduce the engineering effort required for writing test applications, demonstrating the potential of combining static analysis with LLMs to enhance the efficiency of exploit generation and vulnerability validation. Evaluation on Android 12 and 14 systems from Samsung and OnePlus demonstrates PathSentinel's effectiveness, uncovering 51 previously unknown vulnerabilities among 217 apps with only 2 false positives. These results underscore the importance of combining program and access control policy analysis for accurate vulnerability detection and highlight the promising direction of integrating LLMs for automated exploit generation, providing a comprehensive approach to enhancing the security of Android systems against filesystem vulnerabilities.

Aims. Understanding the rotational dynamics of interstellar dust grains is quintessential for the analysis of the observed dust polarization signal. We aim to constrain the set of parameters for an accurate description of the rotational spin-up process of ballistic dust grain aggregates driven by radiative torques (RATs). Methods. We model the dust grains as complex fractal aggregates grown by the ballistic aggregation of uniform spherical particles (monomers) of different sizes. A broad variation of dust materials, shapes, and sizes are studied in the presence of different radiation sources. Results. We find that the canonical parametrization for the torque efficiency overestimates the maximal angular velocity $\omega_{\mathrm{RAT}}$ caused by RATs of grain aggregates, and to resolve this problem we propose a new parametrization which predicts $\omega_{\mathrm{RAT}}$ more accurately. We find that RATs are strongest for larger grains with smaller monomer-density, this manifests as size and monomer-density dependence in the constant part of the parametrization. Following the constant part, the parametrization has two power-laws with different slopes, which retain the universality for all grain sizes. The maximum grain rotation does not scale linearly with radiation strength, due to different drag mechanisms dominating depending on grain material and environment. The angular velocity $\omega_{\mathrm{RAT}}$ of individual single dust grains has a wide distribution and may even differ from the mean by up to two orders of magnitude., Comment: 15 pages, 12 figures

We extend the recently presented, fully exclusive, next-to-leading-order accurate event generator for the simulation of massless neutral- and charged-current deep inelastic scattering (DIS) to the case of incoming neutrinos. The generator can be used to study neutrino-nucleon interactions at (ultra) high energies, and is relevant for a range of fixed-target collider experiments and large-volume neutrino detectors, investigating atmospheric and astrophysical neutrinos. The matching with multi-purpose event generators such as PYTHIA 8 is performed with the POWHEG method, and accounts for parton showering and non-perturbative effects such as hadronization. This makes it possible to investigate higher-order perturbative corrections to realistic observables, such as the distribution of charged particles. To illustrate the capabilities of the code we provide predictions for several differential distributions in fixed-target collisions for neutrino energies up to 1 PeV., Comment: 16 pages. Code available from https://powhegbox.mib.infn.it/

A high precision measurement of the branching fraction of the decay $J/\psi \to p \bar{p} \eta$ is performed using $(10 087 \pm 44) \times 10^6$ $J/\psi$ events recorded by the {BESIII} detector at the {BEPCII} storage ring. The branching fractions of the two decays $J/\psi \to p \bar{p} \eta(\eta \to \gamma\gamma)$ and $J/\psi \to p \bar{p} \eta(\eta \to \pi^+ \pi^- \pi^0)$ are measured individually to be $\mathcal{B}(J/\psi \to p \bar{p} \eta(\eta \to \gamma\gamma)) = (1.480 \pm 0.001 \pm 0.024)\times\,10^{-3}$ and $\mathcal{B}(J/\psi \to p \bar{p} \eta(\eta \to \pi^+ \pi^- \pi^0)) = (1.557 \pm 0.003 \pm 0.038)\times\,10^{-3}$, where the first uncertainties are statistical and the second systematic. Both results are compatible within their uncorrelated systematic uncertainties. The combined result is $\mathcal{B}(J/\psi \to p \bar{p} \eta)=(1.495 \pm 0.001 \pm 0.023)\times\,10^{-3}$ where the first uncertainty is the combined statistical uncertainty and the second one the combined systematic uncertainty of both analyses, incorporating correlations between them. In addition, the $p \bar{p}$ threshold region is investigated for a potential threshold enhancement, and no evidence for one is observed.

Selective Classification, wherein models can reject low-confidence predictions, promises reliable translation of machine-learning based classification systems to real-world scenarios such as clinical diagnostics. While current evaluation of these systems typically assumes fixed working points based on pre-defined rejection thresholds, methodological progress requires benchmarking the general performance of systems akin to the $\mathrm{AUROC}$ in standard classification. In this work, we define 5 requirements for multi-threshold metrics in selective classification regarding task alignment, interpretability, and flexibility, and show how current approaches fail to meet them. We propose the Area under the Generalized Risk Coverage curve ($\mathrm{AUGRC}$), which meets all requirements and can be directly interpreted as the average risk of undetected failures. We empirically demonstrate the relevance of $\mathrm{AUGRC}$ on a comprehensive benchmark spanning 6 data sets and 13 confidence scoring functions. We find that the proposed metric substantially changes metric rankings on 5 out of the 6 data sets.

Using $(27.12\pm 0.14)\times10^8$ $\psi(3686)$ decays and data samples of $e^+e^-$ collisions with $\sqrt{s}$ from 4.130 to 4.780~GeV collected with the BESIII detector, we report the first observation of the electromagnetic Dalitz transition $h_c\to e^+e^-\eta_c$ with a statistical significance of $5.4\sigma$. We measure the ratio of the branching fractions $\frac{\mathcal{B}(h_c\rightarrow e^+e^-\eta_c)}{\mathcal{B}(h_c\rightarrow \gamma \eta_c)}$ separately for the $h_c$ samples produced via $\psi(3686)\to\pi^0h_c$ and $e^+e^-\to\pi^+\pi^-h_c$. The average ratio is determined to be $(0.59\pm0.10(\text{stat.})\pm0.04(\text{syst.}))\%$, where the uncertainty includes both statistical and systematic components.

Acoustic levitation is frequently used for non-contact manipulation of objects and to study the impact of microgravity on physical and biological processes. While the force field produced by sound pressure lifts particles against gravity (primary acoustic force), multiple levitating objects in the same acoustic cavity interact via forces that arise from scattered sound (secondary acoustic forces). Current experimental techniques for obtaining these force fields are not well-suited for mapping the primary force field at high spatial resolution and cannot directly measure the secondary scattering force. Here we introduce a method that can measure both acoustic forces in situ, including secondary forces in the near-field limit between arbitrarily shaped, closely spaced objects. Operating similarly to an atomic force microscope, the method inserts into the acoustic cavity a suitably shaped probe tip at the end of a long, flexible cantilever and optically detects its deflection. This makes it possible to measure forces with a resolution better than 50 nN, and also to apply stress or strain in a controlled manner to manipulate levitated objects. We demonstrate this by extracting the acoustic potential present in a levitation cavity, directly measuring the acoustic scattering force between two objects, and applying tension to a levitated granular raft of acoustically-bound particles in order to obtain the force-displacement curve for its deformation., Comment: The following article has been accepted for publication in Review of Scientific Instruments

We computationally completely enumerate a number of types of row-column designs up to isotopism, including double, sesqui and triple arrays as known from the literature, and two newly introduced types that we call mono arrays and AO-arrays. We calculate autotopism group sizes for the designs we generate. For larger parameter values, where complete enumeration is not feasible, we generate examples of some of the designs, and generate exhaustive lists of admissible parameters. For some admissible parameter sets, we prove non-existence results. We also give some explicit constructions of sesqui arrays, mono arrays and AO-arrays, and investigate connections to Youden rectangles and binary pseud Youden designs.

Modelling passenger assignments in public transport networks is a fundamental task for city planners, especially when deliberating network infrastructure decisions. A key aspect of a realistic model for passenger assignments is to integrate selfish routing behaviour of passengers on the one hand, and the limited vehicle capacities on the other hand. We formulate a side-constrained user equilibrium model in a schedule-based time-expanded transit network, where passengers are modelled via a continuum of non-atomic agents that want to travel with a fixed start time from a user-specific origin to a destination. An agent's route may comprise several rides along given lines, each using vehicles with hard loading capacities. We give a characterization of (side-constrained) user equilibria via a quasi-variational inequality and prove their existence by generalizing a well-known existence result of Bernstein and Smith (Transp. Sci., 1994). We further derive a polynomial time algorithm for single-commodity instances and an exact finite time algorithm for the multi-commodity case. Based on our quasi-variational characterization, we finally devise a fast heuristic computing user equilibria, which is tested on real-world instances based on data gained from the Hamburg S-Bahn system and the Swiss long-distance train network. It turns out that w.r.t. the total travel time, the computed user-equilibria are quite efficient compared to a system optimum, which neglects equilibrium constraints and only minimizes total travel time., Comment: 32 pages, 9 figures

Agent-based models (ABMs) offer a powerful framework for understanding complex systems. However, their computational demands often become a significant barrier as the number of agents and complexity of the simulation increase. Traditional ABM platforms often struggle to fully exploit modern computing resources, hindering the development of large-scale simulations. This paper presents Vahana.jl, a high performance computing open source framework that aims to address these limitations. Building on the formalism of synchronous graph dynamical systems, Vahana.jl is especially well suited for models with a focus on (social) networks. The framework seamlessly supports distribution across multiple compute nodes, enabling simulations that would otherwise be beyond the capabilities of a single machine. Implemented in Julia, Vahana.jl leverages the interactive Read-Eval-Print Loop (REPL) environment, facilitating rapid model development and experimentation.

Recent quench experiments on ultra cold fermions in optical cavities provide a clean platform for studying how long-range interactions between fermions structure their dynamics. Motivated by these experiments, we provide a theoretical analysis of the dynamical instabilities that lead to the formation of superradiance as the hybrid system is driven across the self-organization transition. We compute the rate at which order forms and quantify the fluctuations of the pre-quench state which seed the instability. Our results quantitatively match existing experiments on free fermions and make predictions for quench experiments involving near unitary fermi gases coupled to an optical cavity. Our work suggests that the non-local nature of the photon-mediated interactions between fermions generates ordering dynamics that are qualitatively different than those observed in short-range interacting systems.

The unprecedented statistics of detected Type Ia supernovae (SNe Ia) brought by the Zwicky Transient Facility enables us to probe the impact of the Large-Scale Structure on the properties of these objects. The goal of this paper is to explore the possible impact of the under-dense part of the large-scale structure on the intrinsic SALT2 light curve properties of SNe Ia and uncover possible biases in SN Ia analyses. With a volume-limited selection of ZTF-Cosmo-DR2 Type Ia supernovae overlapping with the SDSS-DR7 survey footprint, we investigate the distribution of their properties with regard to voids detected in the SDSS-DR7 galaxy sample. We further use Voronoi volumes as proxy for local density environments within the large-scale structure. We find a moderate dependency of the stretch toward the localisation around the void centre and none when considering colour. The local Voronoi volumes mostly affect the fraction of low/high stretch supernovae. With the current statistics available, we consider that the impact of high or low local density environment can be considered as a proxy for the colour of the host galaxy. Under-dense environments should not cause any biases in supernova analyses., Comment: 10 pages, 8 figures. Submitted to A&A

Predominant methods for image-based drone detection frequently rely on employing generic object detection algorithms like YOLOv5. While proficient in identifying drones against homogeneous backgrounds, these algorithms often struggle in complex, highly textured environments. In such scenarios, drones seamlessly integrate into the background, creating camouflage effects that adversely affect the detection quality. To address this issue, we introduce a novel deep learning architecture called YOLO-FEDER FusionNet. Unlike conventional approaches, YOLO-FEDER FusionNet combines generic object detection methods with the specialized strength of camouflage object detection techniques to enhance drone detection capabilities. Comprehensive evaluations of YOLO-FEDER FusionNet show the efficiency of the proposed model and demonstrate substantial improvements in both reducing missed detections and false alarms., Comment: 7 pages, 4 figures, 6 tables, to be published in the conference proceedings of the 2024 IEEE International Conference on Image Processing (ICIP)

We study the dynamics of a seeding experiment where a fibrous scaffold material is colonized by two types of cell populations. The specific application that we have in mind is related to the idea of meniscus tissue regeneration. In order to support the development of a promising replacement material, we discuss certain rate equations for the densities of human mesenchymal stem cells and chondrocytes and for the production of collagen-containing extracellular matrix. For qualitative studies, we start with a system of ordinary differential equations and refine then the model to include spatial effects of the underlying nonwoven scaffold structure. Numerical experiments as well as a complete set of parameters for future benchmarking are provided., Comment: 6 pages, 5 figures, 1 table

Control of frictional interactions among liquid-suspended particles has led to tunable, strikingly non-Newtonian rheology via the formation of strong flow constraints as particles come into close proximity under shear. Typically, these frictional interactions have been in the form of physical contact, controllable via particle shape and surface roughness. We investigate a different route, where molecular bridging between nearby particle surfaces generates a controllable "sticky" friction. This is achieved with surface-functionalized colloidal particles capable of forming dynamic covalent bonds with telechelic polymers that comprise the suspending fluid. At low shear stress this results in particles coated with a uniform polymer brush layer. Beyond an onset stress the telechelic polymers become capable of bridging and generate shear thickening. Over the size range investigated, we find that the dynamic brush layer leads to dependence of the onset stress on particle diameter that closely follows a power law with exponent -1.76. In the shear thickening regime, we observe an enhanced dilation in measurements of the first normal stress difference and reduction in the extrapolated volume fraction required for jamming, both consistent with an effective particle friction that increases with decreasing particle diameter. These results are discussed in light of predictions for suspensions of hard spheres and of polymer-grafted particles.

Type Ia supernova (SNIa) are excellent probes of local distance, and the increasing sample sizes of SNIa have driven an increased need to study the associated systematic uncertainties and improve the standardisation methods in preparation for the next generation of cosmological surveys into the dark energy equation-of-state $w$. We aim to probe the potential change in the SNIa standardisation parameter $c$ with redshift and the host-galaxy of the supernova. Improving the standardisation of SNIa brightnesses will require accounting for the relationship between the host and the SNIa, and potential shifts in the SNIa standardisation parameters with redshift will cause biases in the recovered cosmology. Here, we assemble a volume-limited sample of ~3000 likely SNIa across a redshift range of $z = 0.015$ to $z = 0.36$. This sample is fitted with changing mass and redshift bins to determine the relationship between intrinsic properties of SNe Ia and their redshift and host galaxy parameters. We then investigate the colour-luminosity parameter $\beta$ as a further test of the SNIa standardisation process. We find that the changing colour distribution of SNe Ia with redshift is driven by dust at a confidence of $>4\sigma$. Additionally, we show a strong correlation between the host galaxy mass and the colour-luminosity coefficient $\beta$ ($> 4\sigma$), even when accounting for the quantity of dust in a host galaxy.

To date, most investigations on surprisal and entropy effects in reading have been conducted on the group level, disregarding individual differences. In this work, we revisit the predictive power of surprisal and entropy measures estimated from a range of language models (LMs) on data of human reading times as a measure of processing effort by incorporating information of language users' cognitive capacities. To do so, we assess the predictive power of surprisal and entropy estimated from generative LMs on reading data obtained from individuals who also completed a wide range of psychometric tests. Specifically, we investigate if modulating surprisal and entropy relative to cognitive scores increases prediction accuracy of reading times, and we examine whether LMs exhibit systematic biases in the prediction of reading times for cognitively high- or low-performing groups, revealing what type of psycholinguistic subject a given LM emulates. Our study finds that in most cases, incorporating cognitive capacities increases predictive power of surprisal and entropy on reading times, and that generally, high performance in the psychometric tests is associated with lower sensitivity to predictability effects. Finally, our results suggest that the analyzed LMs emulate readers with lower verbal intelligence, suggesting that for a given target group (i.e., individuals with high verbal intelligence), these LMs provide less accurate predictability estimates., Comment: ACL 2024 Findings

Semantic segmentation is an essential component of medical image analysis research, with recent deep learning algorithms offering out-of-the-box applicability across diverse datasets. Despite these advancements, segmentation failures remain a significant concern for real-world clinical applications, necessitating reliable detection mechanisms. This paper introduces a comprehensive benchmarking framework aimed at evaluating failure detection methodologies within medical image segmentation. Through our analysis, we identify the strengths and limitations of current failure detection metrics, advocating for the risk-coverage analysis as a holistic evaluation approach. Utilizing a collective dataset comprising five public 3D medical image collections, we assess the efficacy of various failure detection strategies under realistic test-time distribution shifts. Our findings highlight the importance of pixel confidence aggregation and we observe superior performance of the pairwise Dice score (Roy et al., 2019) between ensemble predictions, positioning it as a simple and robust baseline for failure detection in medical image segmentation. To promote ongoing research, we make the benchmarking framework available to the community., Comment: This work has been submitted for possible publication

Type Ia supernova (SN Ia) cosmology relies on the estimation of lightcurve parameters to derive precision distances that leads to the estimation of cosmological parameters. The empirical SALT2 lightcurve modeling that relies on only two parameters, a stretch x1, and a color c, has been used by the community for almost two decades. In this paper we study the ability of the SALT2 model to fit the nearly 3000 cosmology-grade SN Ia lightcurves from the second release of the Zwicky Transient Facility (ZTF) cosmology science working group. While the ZTF data was not used to train SALT2, the algorithm is modeling the ZTF SN Ia optical lightcurves remarkably well, except for lightcurve points prior to -10 d from maximum, where the training critically lacks statistics. We find that the lightcurve fitting is robust against the considered choice of phase-range, but we show the [-10; +40] d range to be optimal in terms of statistics and accuracy. We do not detect any significant features in the lightcurve fit residuals that could be connected to the host environment. Potential systematic population differences related to the SN Ia host properties might thus not be accountable for by the addition of extra lightcurve parameters. However, a small but significant inconsistency between residuals of blue- and red-SN Ia strongly suggests the existence of a phase-dependent color term, with potential implications for the use of SNe Ia in precision cosmology. We thus encourage modellers to explore this avenue and we emphasize the importance that SN Ia cosmology must include a SALT2 retraining to accurately model the lightcurves and avoid biasing the derivation of cosmological parameters., Comment: 10 pages, 9 figures. Submitted to Astronomy and Astrophysics