Search

Contrastive pretraining can substantially increase model generalisation and downstream performance. However, the quality of the learned representations is highly dependent on the data augmentation strategy applied to generate positive pairs. Positive contrastive pairs should preserve semantic meaning while discarding unwanted variations related to the data acquisition domain. Traditional contrastive pipelines attempt to simulate domain shifts through pre-defined generic image transformations. However, these do not always mimic realistic and relevant domain variations for medical imaging such as scanner differences. To tackle this issue, we herein introduce counterfactual contrastive learning, a novel framework leveraging recent advances in causal image synthesis to create contrastive positive pairs that faithfully capture relevant domain variations. Our method, evaluated across five datasets encompassing both chest radiography and mammography data, for two established contrastive objectives (SimCLR and DINO-v2), outperforms standard contrastive learning in terms of robustness to acquisition shift. Notably, counterfactual contrastive learning achieves superior downstream performance on both in-distribution and on external datasets, especially for images acquired with scanners under-represented in the training set. Further experiments show that the proposed framework extends beyond acquisition shifts, with models trained with counterfactual contrastive learning substantially improving subgroup performance across biological sex., Comment: Code available at https://github.com/biomedia-mira/counterfactual-contrastive/

We demonstrate that two-dimensional Kramers-Weyl fermions can be engineered in spin-orbit coupled twisted bilayers, where the chiral structure of these moir\'e systems breaks all mirror symmetries, confining Kramers-Weyl fermions to high-symmetry points in the Brillouin zone under time reversal symmetry. Our theoretical analysis reveals a symmetry-enforced Weyl-like interlayer moir\'e coupling that universally ensures an ideal radial spin-texture at arbitrary twist angles, under $C_{nz}$ symmetry with n>2. First principles density functional calculation confirm the realization of these fermions in twisted $\alpha$-In$_2$Se$_3$ bilayers, where flat bands and out-of-plane ferroelectric polarization in each layer guarantee two-dimensional Kramers-Weyl physics with perfectly ideal radial spin textures.

The issue of consistency is crucial in quantum gravity. It has recently been intensively addressed for effective symmetry-reduced models. In this article, we exhaustively study the anomaly freedom of effective loop quantum cosmology with generalized holonomy corrections, considering loop correction of the constraints at the perturbative order. We pedagogically explain why, although the holonomy correction -- including the details of the chosen scheme -- applied on the background part of the constraints is crucial, it becomes irrelevant when implemented on perturbative expansions, in the sense that all consequences are "absorbed" in the counter-terms used for the regularization. The possibility of closing the algebra of constraints without counter-terms is also studied. It is argued that, although enforcing a first-class algebra is a strong requirement, this can be achieved in several different ways, often overlooked, which generates ambiguities on the restriction of the form of the generalized holonomy correction. Those ambiguities are examined in details, leading to the conclusion that the consistency of the effective theory for cosmological perturbations, especially when considering scalar modes, cannot be achieved without counter-terms. We also take the opportunity of this work to clarify, as much as possible, all the required steps so that future works have a clear material at disposal. In particular, a highly detailed calculation of all the brackets is provided, emphasizing the (usually implicit) assumptions, hypotheses and manipulations required to ensure the closure of the algebra. Prospects for future works are underlined., Comment: 45 pages, no figures. Any questions, comments, and criticism are welcome

The number of samples in structural brain MRI studies is often too small to properly train deep learning models. Generative models show promise in addressing this issue by effectively learning the data distribution and generating high-fidelity MRI. However, they struggle to produce diverse, high-quality data outside the distribution defined by the training data. One way to address the issue is using causal models developed for 3D volume counterfactuals. However, accurately modeling causality in high-dimensional spaces is a challenge so that these models generally generate 3D brain MRIS of lower quality. To address these challenges, we propose a two-stage method that constructs a Structural Causal Model (SCM) within the latent space. In the first stage, we employ a VQ-VAE to learn a compact embedding of the MRI volume. Subsequently, we integrate our causal model into this latent space and execute a three-step counterfactual procedure using a closed-form Generalized Linear Model (GLM). Our experiments conducted on real-world high-resolution MRI data (1mm) demonstrate that our method can generate high-quality 3D MRI counterfactuals.

Nowadays, deep neural networks are used for solving complex tasks in several critical applications and protecting both their integrity and intellectual property rights (IPR) has become of utmost importance. To this end, we advance WaterMAS, a substitutive, white-box neural network watermarking method that improves the trade-off among robustness, imperceptibility, and computational complexity, while making provisions for increased data payload and security. WasterMAS insertion keeps unchanged the watermarked weights while sharpening their underlying gradient space. The robustness is thus ensured by limiting the attack's strength: even small alterations of the watermarked weights would impact the model's performance. The imperceptibility is ensured by inserting the watermark during the training process. The relationship among the WaterMAS data payload, imperceptibility, and robustness properties is discussed. The secret key is represented by the positions of the weights conveying the watermark, randomly chosen through multiple layers of the model. The security is evaluated by investigating the case in which an attacker would intercept the key. The experimental validations consider 5 models and 2 tasks (VGG16, ResNet18, MobileNetV3, SwinT for CIFAR10 image classification, and DeepLabV3 for Cityscapes image segmentation) as well as 4 types of attacks (Gaussian noise addition, pruning, fine-tuning, and quantization). The code will be released open-source upon acceptance of the article.

Although fundamental to the advancement of Machine Learning, the classic evaluation metrics extracted from the confusion matrix, such as precision and F1, are limited. Such metrics only offer a quantitative view of the models' performance, without considering the complexity of the data or the quality of the hit. To overcome these limitations, recent research has introduced the use of psychometric metrics such as Item Response Theory (IRT), which allows an assessment at the level of latent characteristics of instances. This work investigates how IRT concepts can enrich a confusion matrix in order to identify which model is the most appropriate among options with similar performance. In the study carried out, IRT does not replace, but complements classical metrics by offering a new layer of evaluation and observation of the fine behavior of models in specific instances. It was also observed that there is 97% confidence that the score from the IRT has different contributions from 66% of the classical metrics analyzed., Comment: 15 pages, 8 figures, 3 tables, submitted for the BRACIS'24 conference

Complexity science, despite its broad scope and potential impact, has not kept pace with fields like artificial intelligence, biotechnology and social sciences in addressing ethical concerns. The field lacks a comprehensive ethical framework, leaving us, as a community, vulnerable to ethical challenges and dilemmas. Other areas have gone through similar experiences and created, with discussions and working groups, their guides, policies and recommendations. Therefore, here we highlight the critical absence of formal guidelines, dedicated ethical committees, and widespread discussions on ethics within the complexity science community. Drawing on insights from the disciplines mentioned earlier, we propose a roadmap to enhance ethical awareness and action. Our recommendations include (i) initiating supportive mechanisms to develop ethical guidelines specific to complex systems research, (ii) creating open-access resources, and (iii) fostering inclusive dialogues to ensure that complexity science can responsibly tackle societal challenges and achieve a more inclusive environment. By initiating this dialogue, we aim to encourage a necessary shift in how ethics is integrated into complexity research, positioning the field to address contemporary challenges more effectively., Comment: 7 pages, 2 figures, 1 Annexus, 1 table

This study investigates vulnerabilities in dependencies of sampled open-source software (OSS) projects, the relationship between these and overall project security, and how developers' behaviors and practices influence their mitigation. Through analysis of OSS projects, we have identified common issues in outdated or unmaintained dependencies, including pointer dereferences and array bounds violations, that pose significant security risks. We have also examined developer responses to formal verifier reports, noting a tendency to dismiss potential issues as false positives, which can lead to overlooked vulnerabilities. Our results suggest that reducing the number of direct dependencies and prioritizing well-established libraries with strong security records are effective strategies for enhancing the software security landscape. Notably, four vulnerabilities were fixed as a result of this study, demonstrating the effectiveness of our mitigation strategies., Comment: 16 pages

Scientific Machine Learning is transforming traditional engineering industries by enhancing the efficiency of existing technologies and accelerating innovation, particularly in modeling chemical reactions. Despite recent advancements, the issue of solving stiff chemically reacting problems within computational fluid dynamics remains a significant issue. In this study we propose a novel approach utilizing a multi-layer-perceptron mixer architecture (MLP-Mixer) to model the time-series of stiff chemical kinetics. We evaluate this method using the ROBER system, a benchmark model in chemical kinetics, to compare its performance with traditional numerical techniques. This study provides insight into the industrial utility of the recently developed MLP-Mixer architecture to model chemical kinetics and provides motivation for such neural architecture to be used as a base for time-series foundation models., Comment: Accepted into the IEEE CAI 2024 Workshop on Scientific Machine Learning and Its Industrial Applications (SMLIA2024)

Geometric frustration is known to completely damage kinetic processes of some of the orbitals (and their associated quantum coherence) as to produce flat bands in the non-interacting systems. The impact of introducing additional interaction to the system in such frustrated systems is, however, a highly controversial issue. On the one hand, numerical studies on geometrically frustrated systems of hard-core boson (equivalent to a spin-1/2 systems) typically lead to glass or solid phases containing only local many-body coherence, indicating the persistence of the damage in quantum coherence. On the other, there continues to be noticeable claims of development of superfluidity that implies kinetic flow of particles. To resolve this apparent contradiction of great significance, we present a rigorous proof showing that density-density interaction is incapable of defeating the geometric frustration to allow propagation of those immobile particles, let alone sustaining a superfluidity. Instead, the frustrated systems develop many $\textit{compact}$ many-body localized states as "many-body scars" that do not thermalize, making them good candidates for storing $\textit{robust}$ and $\textit{manipulable}$ quantum information., Comment: 6 pages, 4 figures

In this work, we derive a generalized constitutive relation describing the current response to external electromagnetic fields in electrically biased quantum materials. While our semiclassical Boltzmann approach reveals the existence of electro-optic effects induced by the Berry curvature dipole of Bloch electrons, we also find a wealth of alternative electro-optic effects originating from the interplay between Berry curvature and magnetic moment. In particular, our symmetry analysis indicates the existence of a magnetoelectric electro-optic effect, derived from the simultaneous presence of Berry curvature and magnetic moment, that requires either time-reversal or inversion symmetry breaking. The revealed conductivity coefficients are explicitly written and we derive the tensor shape describing such alternative electromagnetic responses for chiral materials pertaining to space groups 152 and 198.

The quasiparticle density observed in low-temperature superconducting circuits is several orders of magnitude higher than the value expected at thermal equilibrium. The tunneling of this excess of quasiparticles across Josephson junctions is recognized as one of the main loss and decoherence mechanisms in superconducting qubits. Here we propose an additional loss mechanism arising from nonequilibrium quasiparticle densities: Ohmic loss due to quasiparticles residing in superconducting wires away from the junctions. Our theory leverages the recent experimental demonstration that the excess quasiparticles are in quasiequilibrium [T. Connolly et al., $\href{https://doi.org/10.1103/PhysRevLett.132.217001}{Phys. Rev. Lett. {\bf 132}, 217001 (2024)}$] and uses a generalized fluctuation-dissipation theorem to predict the amount of charge and flux noise generated by them. We show that the resulting charge noise can be larger than dielectric loss due to amorphous two-level systems at frequencies in the MHz range, and find a logarithmic-in-frequency ``nearly white" contribution to flux noise that is comparable to the flux noise observed in experiments. This shows that wire-resident quasiparticles are a universal source of loss and decoherence even when the quasiparticles are far away from Josephson junctions., Comment: 7 pages, 3 figures

With the goal of understanding whether or not it is possible to construct a string theory which is consistent with loop quantum gravity (LQG), we study alternate versions of the Nambu-Goto action for a bosonic string. We consider two types of modifications. The first is a phenomenological action based on the observation that LQG tells us that areas of two-surfaces are operators in quantum geometry and are bounded from below. This leads us to a string action which is similar to that of bimetric gravity. We provide formulations of the bimetric string for both the Nambu-Goto (second order) and Polyakov (first order) formulations. We explore the classical solutions of this action and its quantization and relate it to the conventional string solutions. The second is an action in which the background geometry is described in terms of the pullback of the connection which describes the bulk geometry to the worldsheet. The resulting action is in the form of a gauged sigma model, where the spacetime co-ordinates are now vectors which transform under $ISO(D,1)$. We find that for the particular case of a constant background connection the action reduces to the bimetric action discussed above. We discuss classical solutions and quantization strategies for this action and its implications for the broader program of unifying string theory and loop quantum gravity., Comment: Comments/criticisms very welcome; 23 pages, no figures

The International Muon Collider Collaboration (IMCC) [1] was established in 2020 following the recommendations of the European Strategy for Particle Physics (ESPP) and the implementation of the European Strategy for Particle Physics-Accelerator R&D Roadmap by the Laboratory Directors Group [2], hereinafter referred to as the the European LDG roadmap. The Muon Collider Study (MuC) covers the accelerator complex, detectors and physics for a future muon collider. In 2023, European Commission support was obtained for a design study of a muon collider (MuCol) [3]. This project started on 1st March 2023, with work-packages aligned with the overall muon collider studies. In preparation of and during the 2021-22 U.S. Snowmass process, the muon collider project parameters, technical studies and physics performance studies were performed and presented in great detail. Recently, the P5 panel [4] in the U.S. recommended a muon collider R&D, proposed to join the IMCC and envisages that the U.S. should prepare to host a muon collider, calling this their "muon shot". In the past, the U.S. Muon Accelerator Programme (MAP) [5] has been instrumental in studies of concepts and technologies for a muon collider., Comment: This document summarises the International Muon Collider Collaboration (IMCC) progress and status of the Muon Collider R&D programme

We report a new 40 ppb measurement of the positronium $\text{1}^\text{3}\text{S}_\text{1} \to \text{2}^\text{3}\text{S}_\text{1}$ interval using pulsed two-photon optical spectroscopy. The transition is detected via field-ionization of atoms excited from the 2S to the 20P Rydberg state. Precise Monte-Carlo line-shape simulations allow for the accounting of effects such as Doppler and AC Stark shifts, while an optical heterodyne measurement of the excitation laser pulse is used to correct for laser frequency chirp. A value of $1\,233\,607\,210.5\pm 49.6\, \mathrm{MHz}$ was obtained. This scheme allows for the measurement of the velocity distribution of positronium atoms to correct for the second-order Doppler effect. This is the major source of systematic uncertainty expected for future measurements of this transition with a CW laser, thus, our technique paves the way toward a new generation of a high precision determination of this interval in positronium., Comment: 20 pages, 29 figures

Software engineers have adopted the Dynamic Software Product Lines (DSPL) engineering practices to develop Dynamically Adaptable Software (DAS). DAS is seen as a DSPL application and must cope with a large number of configurations of features, Non-functional Requirements (NFRs), and contexts. However, the accurate representation of the impact of features over NFRs and contexts for the identification of optimal configurations is not a trivial task. Software engineers need to have domain knowledge and design DAS before deploying to satisfy those requirements. Aiming to handle them, we proposed an approach of Trade-off Analysis for DSPL at design-time, named ToffA-DSPL. It deals with the configuration selection process considering interactions between NFRs and contexts. We performed an exploratory study based on simulations to identify the usefulness of the ToffA-DSPL approach. In general, the configurations suggested by ToffA-DSPL provide high satisfaction levels of NFRs. Based on simulations, we evidenced that our approach aims to explore reuse and is useful for generating valid and optimal configurations. In addition, ToffA-DSPL enables software engineers to conduct trade-off analysis, evaluate changes in the context feature, and define an adaptation model from optimal configurations found in the analysis.

Learning modular object-centric representations is crucial for systematic generalization. Existing methods show promising object-binding capabilities empirically, but theoretical identifiability guarantees remain relatively underdeveloped. Understanding when object-centric representations can theoretically be identified is crucial for scaling slot-based methods to high-dimensional images with correctness guarantees. To that end, we propose a probabilistic slot-attention algorithm that imposes an aggregate mixture prior over object-centric slot representations, thereby providing slot identifiability guarantees without supervision, up to an equivalence relation. We provide empirical verification of our theoretical identifiability result using both simple 2-dimensional data and high-resolution imaging datasets.

We construct 1-parameter families of well-known solutions to the Yamabe problem defined on Aloff-Wallach Spaces to determine bifurcation instants for these homogeneous spaces by examining changes in the Morse index of these metrics as the parameter varies over the positive real numbers. A bifurcation point for such families is an accumulation point of other solutions to the Yamabe problem, while a local rigidity point is an isolated solution of this problem, i.e., it is not a bifurcation point.

In scenarios coined Maxwell's demon, information on microscopic degrees of freedom is used to seemingly violate the second law of thermodynamics. This has been studied in the classical as well as the quantum domain. In this paper, we study an implementation of Maxwell's demon that can operate in both domains. In particular, we investigate information-to-work conversion over the quantum-to-classical transition. The demon continuously measures the charge state of a double quantum dot, and uses this information to guide electrons against a voltage bias by tuning the on-site energies of the dots. Coherent tunneling between the dots allows for the buildup of quantum coherence in the system. Under strong measurements, the coherence is suppressed, and the system is well-described by a classical model. As the measurement strength is further increased, the Zeno effect prohibits interdot tunneling. A Zeno-like effect is also observed for weak measurements, where measurement errors lead to fluctuations in the on-site energies, dephasing the system. We anticipate similar behaviors in other quantum systems under continuous measurement and feedback control, making our results relevant for implementations in quantum technology and quantum control.

Advances in material fabrication techniques and growth methods have opened up a new chapter for twistronics, in the form of twisted freestanding three-dimensional material membranes. Through first-principles calculations based on density functional theory, we investigate the crystal and electronic structures of twisted bilayer BaTiO$_3$. Our findings reveal that large stacking fault energy leads to chiral in-plane vortex pattern that was recently observed in experiments. Moreover, we also found non-zero out-of-plane local dipole moments, indicating that the strong interlayer interaction might offer promising strategy to stabilize ferroelectric order in the two-dimensional limit. Remarkably, the vortex pattern in the twisted BaTiO$_3$ bilayer support localized electronic states with quasi-flat bands, associated with the interlayer hybridization of oxygen $p_z$ orbitals. We found that the associated band width reaches a minimum at $\sim$19$^{\circ}$ twisting, configuring the largest magic angle in moir\'e systems reported so far. Further, the moir\'e vortex pattern bears a striking resemblance to two interpenetrating Lieb lattices and corresponding tight-binding model provides a comprehensive description of the evolution the moir\'e bands with twist angle and reveals the topological nature of these states., Comment: 8 pages, 5 figures

In a typical semiconductor material, the majority of heat is carried by long wavelength, long mean-free-path phonons. Nanostructuring strategies to reduce thermal conductivity, a promising direction in the field of thermoelectrics, place scattering centers of size and spatial separation comparable to the mean-free-paths of the dominant phonons to selectively scatter them. The resultant thermal conductivity is in most cases well predicted using Matthiessens rule. In general, however, long wavelength phonons are not as effectively scattered as the rest of the phonon spectrum. In this work, using large-scale Molecular Dynamics simulations, Non-Equilibrium Greens Function simulations, and Monte Carlo simulations, we show that specific nanoporous geometries, which create narrow constrictions in the passage of phonons, lead to anticorrelated heat currents in the phonon spectrum. This results in super-suppression of long-wavelength phonons due to heat trapping, and reductions in the thermal conductivity well below what is predicted by Matthiessens rule., Comment: 25 pages, 7 figures

Quantum technologies and experiments often require preparing systems in low-temperature states. Here, we investigate cooling schemes using feedback protocols modeled with a Quantum Fokker-Planck Master Equation (QFPME) recently derived by Annby-Andersson et. al. (Phys. Rev. Lett. 129, 050401, 2022). This equation describes systems under continuous weak measurements, with feedback based on the outcome of these measurements. We apply this formalism to study the cooling and trapping of a harmonic oscillator for several protocols based on position and/or momentum measurements. We find that the protocols can cool the oscillator down to, or close to, the ground state for suitable choices of parameters. Our analysis provides an analytically solvable case study of quantum measurement and feedback and illustrates the application of the QFPME to continuous quantum systems., Comment: 18 pages, 3 figures

The recently introduced class of architectures known as Neural Operators has emerged as highly versatile tools applicable to a wide range of tasks in the field of Scientific Machine Learning (SciML), including data representation and forecasting. In this study, we investigate the capabilities of Neural Implicit Flow (NIF), a recently developed mesh-agnostic neural operator, for representing the latent dynamics of canonical systems such as the Kuramoto-Sivashinsky (KS), forced Korteweg-de Vries (fKdV), and Sine-Gordon (SG) equations, as well as for extracting dynamically relevant information from them. Finally we assess the applicability of NIF as a dimensionality reduction algorithm and conduct a comparative analysis with another widely recognized family of neural operators, known as Deep Operator Networks (DeepONets)., Comment: Accepted into the International conference on Scientific Computation and Machine Learning 2024 (SCML 2024)

The inquiry into the origin of brain complexity remains a pivotal question in neuroscience. While synaptic stimuli are acknowledged as significant, their efficacy often falls short in elucidating the extensive interconnections of the brain and nuanced levels of cognitive integration. Recent advances in neuroscience have brought the mechanisms underlying the generation of highly intricate dynamics, emergent patterns, and sophisticated oscillatory signals into question. Within this context, our study, in alignment with current research, posits the hypothesis that ephaptic communication may emerge as the primary candidate for unraveling optimal brain complexity. In this investigation, we conducted a comparative analysis between two types of networks utilizing the Quadratic Integrate-and-Fire Ephaptic model (QIF-E): (I) a small-world synaptic network (ephaptic-off) and (II) a mixed composite network comprising a small-world synaptic network with the addition of an ephaptic network (ephaptic-on). Utilizing the Multiscale Entropy methodology, we conducted an in-depth analysis of the responses generated by both network configurations, with complexity assessed by integrating across all temporal scales. Our findings demonstrate that ephaptic coupling enhances complexity under specific topological conditions, considering variables such as time, spatial scales, and synaptic intensity. These results offer fresh insights into the dynamics of communication within the nervous system and underscore the fundamental role of ephapticity in regulating complex brain functions.

Cyclic classical and quantum thermal machines show higher efficiency when the strokes are carried out quasi-statically. Recent theoretical and experimental work on figures of merit for thermal machines show that they have an advantage when operating in environments with negative temperatures. In an experimental proof of concept [Phys. Rev. Lett. 122, 240602 (2019)], it was shown that quantum Otto engines operating at negative temperatures can exhibit a behavior in which the faster the cycle is carried out, the higher the efficiency. In this work, we make use of the concept of entropy production and friction work to explain this counterintuitive behavior, and we show that it only occurs when reservoirs have negative temperatures., Comment: 6 pages, 5 figures

Two different methods are used to study the existence and stability of the (1+1)-dimensional $\Phi^4$ oscillon. The variational technique approximates it by a periodic function with a set of adiabatically changing parameters. An alternative approach treats oscillons as standing waves in a finite-size box; these are sought as solutions of a boundary-value problem on a two-dimensional domain. The numerical analysis reveals that the standing wave's energy-frequency diagram is fragmented into disjoint segments with $\omega_{n+1} < \omega < \omega_{n}$, where $\omega_n= \omega_0/ (n+1)$, $n=0,1,2, ...$, and $\omega_0$ is the endpoint of the continuous spectrum (mass threshold of the model). The variational approximation involving the first, zeroth and second harmonic components provides an accurate description of the oscillon with the frequency in $(\omega_1, \omega_0)$, but breaks down as $\omega$ falls out of that interval., Comment: 24 pages, 7 figures. Presented at XLIV Dynamics Days Europe (Constructor University, Bremen, July 29 - August 2, 2024)

Established in 2012 by members of the Food and Agriculture Organisation (FAO), the Global Soil Partnership (GSP) is a global network of stakeholders promoting sound land and soil management practices towards a sustainable world food system. However, soil survey largely remains a local or regional activity, bound to heterogeneous methods and conventions. Recognising the relevance of global and trans-national policies towards sustainable land management practices, the GSP elected data harmonisation and exchange as one of its key lines of action. Building upon international standards and previous work towards a global soil data ontology, an improved domain model was eventually developed within the GSP [54], the basis for a Global Soil Information System (GloSIS). This work also identified the Semantic Web as a possible avenue to operationalise the domain model. This article presents the GloSIS web ontology, an implementation of the GloSIS domain model with the Web Ontology Language (OWL). Thoroughly employing a host of Semantic Web standards (SOSA, SKOS, GeoSPARQL, QUDT), GloSIS lays out not only a soil data ontology but also an extensive set of ready-to-use code-lists for soil description and physio-chemical analysis. Various examples are provided on the provision and use of GloSIS-compliant linked data, showcasing the contribution of this ontology to the discovery, exploration, integration and access of soil data.

Causal generative modelling is gaining interest in medical imaging due to its ability to answer interventional and counterfactual queries. Most work focuses on generating counterfactual images that look plausible, using auxiliary classifiers to enforce effectiveness of simulated interventions. We investigate pitfalls in this approach, discovering the issue of attribute amplification, where unrelated attributes are spuriously affected during interventions, leading to biases across protected characteristics and disease status. We show that attribute amplification is caused by the use of hard labels in the counterfactual training process and propose soft counterfactual fine-tuning to mitigate this issue. Our method substantially reduces the amplification effect while maintaining effectiveness of generated images, demonstrated on a large chest X-ray dataset. Our work makes an important advancement towards more faithful and unbiased causal modelling in medical imaging.

Contrastive pretraining is well-known to improve downstream task performance and model generalisation, especially in limited label settings. However, it is sensitive to the choice of augmentation pipeline. Positive pairs should preserve semantic information while destroying domain-specific information. Standard augmentation pipelines emulate domain-specific changes with pre-defined photometric transformations, but what if we could simulate realistic domain changes instead? In this work, we show how to utilise recent progress in counterfactual image generation to this effect. We propose CF-SimCLR, a counterfactual contrastive learning approach which leverages approximate counterfactual inference for positive pair creation. Comprehensive evaluation across five datasets, on chest radiography and mammography, demonstrates that CF-SimCLR substantially improves robustness to acquisition shift with higher downstream performance on both in- and out-of-distribution data, particularly for domains which are under-represented during training., Comment: Accepted for publication at the MICCAI 2024 Data Engineering in Medical Imaging workshop. Code available at https://github.com/biomedia-mira/counterfactual-contrastive. Extended version of this work available at arXiv:2409.10365

We express the $q$-th Gauss--Bonnet--Chern mass of an immersed submanifold of Euclidean space as a linear combination of two terms: the total $(2q)$-th mean curvature and the integral, over the entire manifold, of the inner product between the $(2q+1)$-th mean curvature vector and the position vector of the immersion. As a consequence, we obtain, for each $q$, a geometric inequality that holds whenever the positive mass theorem (for the $q$-th Gauss--Bonnet--Chern mass) holds.