Search

Diffusion models excel at capturing the natural design spaces of images, molecules, DNA, RNA, and protein sequences. However, rather than merely generating designs that are natural, we often aim to optimize downstream reward functions while preserving the naturalness of these design spaces. Existing methods for achieving this goal often require ``differentiable'' proxy models (\textit{e.g.}, classifier guidance or DPS) or involve computationally expensive fine-tuning of diffusion models (\textit{e.g.}, classifier-free guidance, RL-based fine-tuning). In our work, we propose a new method to address these challenges. Our algorithm is an iterative sampling method that integrates soft value functions, which looks ahead to how intermediate noisy states lead to high rewards in the future, into the standard inference procedure of pre-trained diffusion models. Notably, our approach avoids fine-tuning generative models and eliminates the need to construct differentiable models. This enables us to (1) directly utilize non-differentiable features/reward feedback, commonly used in many scientific domains, and (2) apply our method to recent discrete diffusion models in a principled way. Finally, we demonstrate the effectiveness of our algorithm across several domains, including image generation, molecule generation, and DNA/RNA sequence generation. The code is available at \href{https://github.com/masa-ue/SVDD}{https://github.com/masa-ue/SVDD}., Comment: The code is available at https://github.com/masa-ue/SVDD

High-content phenotypic screening, including high-content imaging (HCI), has gained popularity in the last few years for its ability to characterize novel therapeutics without prior knowledge of the protein target. When combined with deep learning techniques to predict and represent molecular-phenotype interactions, these advancements hold the potential to significantly accelerate and enhance drug discovery applications. This work focuses on the novel task of HCI-guided molecular design. Generative models for molecule design could be guided by HCI data, for example with a supervised model that links molecules to phenotypes of interest as a reward function. However, limited labeled data, combined with the high-dimensional readouts, can make training these methods challenging and impractical. We consider an alternative approach in which we leverage an unsupervised multimodal joint embedding to define a latent similarity as a reward for GFlowNets. The proposed model learns to generate new molecules that could produce phenotypic effects similar to those of the given image target, without relying on pre-annotated phenotypic labels. We demonstrate that the proposed method generates molecules with high morphological and structural similarity to the target, increasing the likelihood of similar biological activity, as confirmed by an independent oracle model.

This tutorial provides a comprehensive survey of methods for fine-tuning diffusion models to optimize downstream reward functions. While diffusion models are widely known to provide excellent generative modeling capability, practical applications in domains such as biology require generating samples that maximize some desired metric (e.g., translation efficiency in RNA, docking score in molecules, stability in protein). In these cases, the diffusion model can be optimized not only to generate realistic samples but also to explicitly maximize the measure of interest. Such methods are based on concepts from reinforcement learning (RL). We explain the application of various RL algorithms, including PPO, differentiable optimization, reward-weighted MLE, value-weighted sampling, and path consistency learning, tailored specifically for fine-tuning diffusion models. We aim to explore fundamental aspects such as the strengths and limitations of different RL-based fine-tuning algorithms across various scenarios, the benefits of RL-based fine-tuning compared to non-RL-based approaches, and the formal objectives of RL-based fine-tuning (target distributions). Additionally, we aim to examine their connections with related topics such as classifier guidance, Gflownets, flow-based diffusion models, path integral control theory, and sampling from unnormalized distributions such as MCMC. The code of this tutorial is available at https://github.com/masa-ue/RLfinetuning_Diffusion_Bioseq, Comment: We plan to add more content/codes. Please let us know if there are any comments

Diffusion models are powerful generative models that allow for precise control over the characteristics of the generated samples. While these diffusion models trained on large datasets have achieved success, there is often a need to introduce additional controls in downstream fine-tuning processes, treating these powerful models as pre-trained diffusion models. This work presents a novel method based on reinforcement learning (RL) to add additional controls, leveraging an offline dataset comprising inputs and corresponding labels. We formulate this task as an RL problem, with the classifier learned from the offline dataset and the KL divergence against pre-trained models serving as the reward functions. We introduce our method, $\textbf{CTRL}$ ($\textbf{C}$onditioning pre-$\textbf{T}$rained diffusion models with $\textbf{R}$einforcement $\textbf{L}$earning), which produces soft-optimal policies that maximize the abovementioned reward functions. We formally demonstrate that our method enables sampling from the conditional distribution conditioned on additional controls during inference. Our RL-based approach offers several advantages over existing methods. Compared to commonly used classifier-free guidance, our approach improves sample efficiency, and can greatly simplify offline dataset construction by exploiting conditional independence between the inputs and additional controls. Furthermore, unlike classifier guidance, we avoid the need to train classifiers from intermediate states to additional controls., Comment: Under review

Isolated, undamped geodesic-acoustic-mode (GAM) packets have been demonstrated to obey a (focusing) nonlinear Schr\"odinger equation (NLSE) [E. Poli, Phys. Plasmas 2021]. This equation predicts susceptibility of GAM packets to the modulational instability (MI). The necessary conditions for this instability are analyzed analytically and numerically using the NLSE model. The predictions of the NLSE are compared to gyrokinetic simulations performed with the global particle-in-cell code ORB5, where GAM packets are created from initial perturbations of the axisymmetric radial electric field $E_r$. An instability of the GAM packets with respect to modulations is observed both in cases in which an initial perturbation is imposed and when the instability develops spontaneously. However, significant differences in the dynamics of the small scales are discerned between the NLSE and gyrokinetic simulations. These discrepancies are mainly due to the radial dependence of the strength of the nonlinear term, which we do not retain in the solution of the NLSE, and to the damping of higher spectral components. The damping of the high-$k_r$ components, which develop as a consequence of the nonlinearity, can be understood in terms of Landau damping. The influence of the ion Larmor radius $\rho_i$ as well as the perturbation wavevector $k_{\text{pert}}$ on this effect is studied. For the parameters considered here the aforementioned damping mechanism hinders the MI process significantly from developing to its full extent and is strong enough to stabilize some of the (according to the undamped NLSE model) unstable wavevectors., Comment: 26 pages, 20 figures. This submission is associated to a contribution to the 20th European Fusion Theory Conference (EFTC) in Padua 2023. Submitted to the Journal of Plasma Physics issue for contributions from the 20th European Fusion Theory Conference

AI-driven design problems, such as DNA/protein sequence design, are commonly tackled from two angles: generative modeling, which efficiently captures the feasible design space (e.g., natural images or biological sequences), and model-based optimization, which utilizes reward models for extrapolation. To combine the strengths of both approaches, we adopt a hybrid method that fine-tunes cutting-edge diffusion models by optimizing reward models through RL. Although prior work has explored similar avenues, they primarily focus on scenarios where accurate reward models are accessible. In contrast, we concentrate on an offline setting where a reward model is unknown, and we must learn from static offline datasets, a common scenario in scientific domains. In offline scenarios, existing approaches tend to suffer from overoptimization, as they may be misled by the reward model in out-of-distribution regions. To address this, we introduce a conservative fine-tuning approach, BRAID, by optimizing a conservative reward model, which includes additional penalization outside of offline data distributions. Through empirical and theoretical analysis, we demonstrate the capability of our approach to outperform the best designs in offline data, leveraging the extrapolation capabilities of reward models while avoiding the generation of invalid designs through pre-trained diffusion models., Comment: Under review

Diffusion models excel at modeling complex data distributions, including those of images, proteins, and small molecules. However, in many cases, our goal is to model parts of the distribution that maximize certain properties: for example, we may want to generate images with high aesthetic quality, or molecules with high bioactivity. It is natural to frame this as a reinforcement learning (RL) problem, in which the objective is to fine-tune a diffusion model to maximize a reward function that corresponds to some property. Even with access to online queries of the ground-truth reward function, efficiently discovering high-reward samples can be challenging: they might have a low probability in the initial distribution, and there might be many infeasible samples that do not even have a well-defined reward (e.g., unnatural images or physically impossible molecules). In this work, we propose a novel reinforcement learning procedure that efficiently explores on the manifold of feasible samples. We present a theoretical analysis providing a regret guarantee, as well as empirical validation across three domains: images, biological sequences, and molecules., Comment: Accepted at ICML 2024

Diffusion models excel at capturing complex data distributions, such as those of natural images and proteins. While diffusion models are trained to represent the distribution in the training dataset, we often are more concerned with other properties, such as the aesthetic quality of the generated images or the functional properties of generated proteins. Diffusion models can be finetuned in a goal-directed way by maximizing the value of some reward function (e.g., the aesthetic quality of an image). However, these approaches may lead to reduced sample diversity, significant deviations from the training data distribution, and even poor sample quality due to the exploitation of an imperfect reward function. The last issue often occurs when the reward function is a learned model meant to approximate a ground-truth "genuine" reward, as is the case in many practical applications. These challenges, collectively termed "reward collapse," pose a substantial obstacle. To address this reward collapse, we frame the finetuning problem as entropy-regularized control against the pretrained diffusion model, i.e., directly optimizing entropy-enhanced rewards with neural SDEs. We present theoretical and empirical evidence that demonstrates our framework is capable of efficiently generating diverse samples with high genuine rewards, mitigating the overoptimization of imperfect reward models., Comment: Under review (codes will be released soon)

In this work, we use the global electromagnetic and electrostatic gyro kinetic approaches to investigate the effects of zonal flows forced-driven by Alfv\'en modes due to their excitation by energetic particles (EPs), on the dynamics of ITG (Ion temperature gradient) instabilities. The equilibrium of the 92416 JET tokamak shot is considered. The linear and nonlinear Alfv\'en modes dynamics, as well as the zonal flow dynamics, are investigated and their respective radial structures and saturation levels are reported. ITG dynamics in the presence of the zonal flows excited by these Alfv\'en modes are also investigated. We find that, the zonal flows forced-driven by Alfv\'en modes can significantly impact the ITG dynamics. A zonal flow amplitude scan reveals the existence of an inverse relation between the zonal flow amplitude and the ITG growth rate. These results show that, forced-driven zonal flows can be an important indirect part of turbulence mitigation due to the injection of energetic particles., Comment: 13 pages, 10 figures

Colonoscopy plays a crucial role in the diagnosis and prognosis of various gastrointestinal diseases. Due to the challenges of collecting large-scale high-quality ground truth annotations for colonoscopy images, and more generally medical images, we explore using self-supervised features from vision transformers in three challenging tasks for colonoscopy images. Our results indicate that image-level features learned from DINO models achieve image classification performance comparable to fully supervised models, and patch-level features contain rich semantic information for object detection. Furthermore, we demonstrate that self-supervised features combined with unsupervised segmentation can be used to discover multiple clinically relevant structures in a fully unsupervised manner, demonstrating the tremendous potential of applying these methods in medical image analysis.

Uncertainty estimation is critical in high-stakes machine learning applications. One effective way to estimate uncertainty is conformal prediction, which can provide predictive inference with statistical coverage guarantees. We present a new conformal regression method, Spline Prediction Intervals via Conformal Estimation (SPICE), that estimates the conditional density using neural-network-parameterized splines. We prove universal approximation and optimality results for SPICE, which are empirically validated by our experiments. SPICE is compatible with two different efficient-to-compute conformal scores, one oracle-optimal for marginal coverage (SPICE-ND) and the other asymptotically optimal for conditional coverage (SPICE-HPD). Results on benchmark datasets demonstrate SPICE-ND models achieve the smallest average prediction set sizes, including average size reductions of nearly 50% for some datasets compared to the next best baseline. SPICE-HPD models achieve the best conditional coverage compared to baselines. The SPICE implementation is made available.

Nonlinear simulations of Alfv\'en modes (AM) driven by energetic particles (EP) in the presence of turbulence are performed with the gyrokinetic particle-in-cell code ORB5. The AMs carry a heat flux, and consequently they nonlinearly modify the plasma temperature profiles. The isolated effect of this modification on the dynamics of turbulence is studied, by means of electrostatic simulations. We find that turbulence is reduced when the profiles relaxed by the AM are used, with respect to the simulation where the unperturbed profiles are used. This is an example of indirect interaction of EPs and turbulence. First, an analytic magnetic equilibrium with circular concentric flux surfaces is considered as a simplified example for this study. Then, an application to an experimentally relevant case of ASDEX Upgrade is discussed.

Diffusion models have achieved state-of-the-art performance in generating many different kinds of data, including images, text, and videos. Despite their success, there has been limited research on how the underlying diffusion process and the final convergent prior can affect generative performance; this research has also been limited to continuous data types and a score-based diffusion framework. To fill this gap, we explore how different discrete diffusion kernels (which converge to different prior distributions) affect the performance of diffusion models for graphs. To this end, we developed a novel formulation of a family of discrete diffusion kernels which are easily adjustable to converge to different Bernoulli priors, and we study the effect of these different kernels on generative performance. We show that the quality of generated graphs is sensitive to the prior used, and that the optimal choice cannot be explained by obvious statistics or metrics, which challenges the intuitions which previous works have suggested.

Macrocyclic peptides are an emerging therapeutic modality, yet computational approaches for accurately sampling their diverse 3D ensembles remain challenging due to their conformational diversity and geometric constraints. Here, we introduce RINGER, a diffusion-based transformer model using a redundant internal coordinate representation that generates three-dimensional conformational ensembles of macrocyclic peptides from their 2D representations. RINGER provides fast backbone and side-chain sampling while respecting key structural invariances of cyclic peptides. Through extensive benchmarking and analysis against gold-standard conformer ensembles of cyclic peptides generated with metadynamics, we demonstrate how RINGER generates both high-quality and diverse geometries at a fraction of the computational cost. Our work lays the foundation for improved sampling of cyclic geometries and the development of geometric learning methods for peptides.

Computational and machine learning approaches to model the conformational landscape of macrocyclic peptides have the potential to enable rational design and optimization. However, accurate, fast, and scalable methods for modeling macrocycle geometries remain elusive. Recent deep learning approaches have significantly accelerated protein structure prediction and the generation of small-molecule conformational ensembles, yet similar progress has not been made for macrocyclic peptides due to their unique properties. Here, we introduce CREMP, a resource generated for the rapid development and evaluation of machine learning models for macrocyclic peptides. CREMP contains 36,198 unique macrocyclic peptides and their high-quality structural ensembles generated using the Conformer-Rotamer Ensemble Sampling Tool (CREST). Altogether, this new dataset contains nearly 31.3 million unique macrocycle geometries, each annotated with energies derived from semi-empirical extended tight-binding (xTB) DFT calculations. Additionally, we include 3,258 macrocycles with reported passive permeability data to couple conformational ensembles to experiment. We anticipate that this dataset will enable the development of machine learning models that can improve peptide design and optimization for novel therapeutics.

Generative flow networks (GFlowNets) are a family of algorithms that learn a generative policy to sample discrete objects $x$ with non-negative reward $R(x)$. Learning objectives guarantee the GFlowNet samples $x$ from the target distribution $p^*(x) \propto R(x)$ when loss is globally minimized over all states or trajectories, but it is unclear how well they perform with practical limits on training resources. We introduce an efficient evaluation strategy to compare the learned sampling distribution to the target reward distribution. As flows can be underdetermined given training data, we clarify the importance of learned flows to generalization and matching $p^*(x)$ in practice. We investigate how to learn better flows, and propose (i) prioritized replay training of high-reward $x$, (ii) relative edge flow policy parametrization, and (iii) a novel guided trajectory balance objective, and show how it can solve a substructure credit assignment problem. We substantially improve sample efficiency on biochemical design tasks., Comment: Accepted to ICML 2023

Since 2005 ESO has been working with its community and industry to develop an extremely large optical/infrared telescope. ESO's Extremely Large Telescope, or ELT for short, is a revolutionary ground-based telescope that will have a 39-metre main mirror and will be the largest visible and infrared light telescope in the world. To address specific topics that are needed for the science operations and calibrations of the telescope, thirteen specific working groups were created to coordinate the effort between ESO, the instrument consortia, and the wider community. We describe here the goals of these working groups as well as their achievements so far., Comment: 8 pages, 4 figures

Diffusion models achieve state-of-the-art performance in generating realistic objects and have been successfully applied to images, text, and videos. Recent work has shown that diffusion can also be defined on graphs, including graph representations of drug-like molecules. Unfortunately, it remains difficult to perform conditional generation on graphs in a way which is interpretable and controllable. In this work, we propose GraphGUIDE, a novel framework for graph generation using diffusion models, where edges in the graph are flipped or set at each discrete time step. We demonstrate GraphGUIDE on several graph datasets, and show that it enables full control over the conditional generation of arbitrary structural properties without relying on predefined labels. Our framework for graph diffusion can have a large impact on the interpretable conditional generation of graphs, including the generation of drug-like molecules with desired properties in a way which is informed by experimental evidence.

Deep graph generative modeling has proven capable of learning the distribution of complex, multi-scale structures characterizing real-world graphs. However, one of the main limitations of existing methods is their large output space, which limits generation scalability and hinders accurate modeling of the underlying distribution. To overcome these limitations, we propose a novel approach that significantly reduces the output space of existing graph generative models. Specifically, starting from the observation that many real-world graphs have low graph bandwidth, we restrict graph bandwidth during training and generation. Our strategy improves both generation scalability and quality without increasing architectural complexity or reducing expressiveness. Our approach is compatible with existing graph generative methods, and we describe its application to both autoregressive and one-shot models. We extensively validate our strategy on synthetic and real datasets, including molecular graphs. Our experiments show that, in addition to improving generation efficiency, our approach consistently improves generation quality and reconstruction accuracy. The implementation is made available., Comment: Accepted at ICML 2023

Learning causal relationships between variables is a well-studied problem in statistics, with many important applications in science. However, modeling real-world systems remain challenging, as most existing algorithms assume that the underlying causal graph is acyclic. While this is a convenient framework for developing theoretical developments about causal reasoning and inference, the underlying modeling assumption is likely to be violated in real systems, because feedback loops are common (e.g., in biological systems). Although a few methods search for cyclic causal models, they usually rely on some form of linearity, which is also limiting, or lack a clear underlying probabilistic model. In this work, we propose a novel framework for learning nonlinear cyclic causal graphical models from interventional data, called NODAGS-Flow. We perform inference via direct likelihood optimization, employing techniques from residual normalizing flows for likelihood estimation. Through synthetic experiments and an application to single-cell high-content perturbation screening data, we show significant performance improvements with our approach compared to state-of-the-art methods with respect to structure recovery and predictive performance.

Class-labeled datasets, particularly those common in scientific domains, are rife with internal structure, yet current class-conditional diffusion models ignore these relationships and implicitly diffuse on all classes in a flat fashion. To leverage this structure, we propose hierarchically branched diffusion models as a novel framework for class-conditional generation. Branched diffusion models rely on the same diffusion process as traditional models, but learn reverse diffusion separately for each branch of a hierarchy. We highlight several advantages of branched diffusion models over the current state-of-the-art methods for class-conditional diffusion, including extension to novel classes in a continual-learning setting, a more sophisticated form of analogy-based conditional generation (i.e. transmutation), and a novel interpretability into the generation process. We extensively evaluate branched diffusion models on several benchmark and large real-world scientific datasets spanning many data modalities.

Molecular shape and geometry dictate key biophysical recognition processes, yet many graph neural networks disregard 3D information for molecular property prediction. Here, we propose a new contrastive-learning procedure for graph neural networks, Molecular Contrastive Learning from Shape Similarity (MolCLaSS), that implicitly learns a three-dimensional representation. Rather than directly encoding or targeting three-dimensional poses, MolCLaSS matches a similarity objective based on Gaussian overlays to learn a meaningful representation of molecular shape. We demonstrate how this framework naturally captures key aspects of three-dimensionality that two-dimensional representations cannot and provides an inductive framework for scaffold hopping.

How much explicit guidance is necessary for conditional diffusion? We consider the problem of conditional sampling using an unconditional diffusion model and limited explicit guidance (e.g., a noised classifier, or a conditional diffusion model) that is restricted to a small number of time steps. We explore a model predictive control (MPC)-like approach to approximate guidance by simulating unconditional diffusion forward, and backpropagating explicit guidance feedback. MPC-approximated guides have high cosine similarity to real guides, even over large simulation distances. Adding MPC steps improves generative quality when explicit guidance is limited to five time steps.

In this work, we revisit the linear gyro-kinetic theory of geodesic acoustic modes (GAMs) and derive a general dispersion relation for an arbitrary equilibrium distribution function of ions. A bi-Maxwellian distribution of ions is then used to study the effects of ion temperature anisotropy on GAM frequency and growth rate. We find that ion temperature anisotropy yields sensible modifications to both the GAM frequency and growth rate as both tend to increase with anisotropy and these results are strongly affected by the electron to ion temperature ratio.

Kinetic Inductance Detectors (KIDs) are superconducting energy-resolving detectors, sensitive to single photons from the near-infrared to ultraviolet. We study a hybrid KID design consisting of a beta phase tantalum ($\beta$-Ta) inductor and a NbTiN interdigitated capacitor (IDC). The devices show an average intrinsic quality factor $Q_i$ of 4.3$\times10^5$ $\pm$ 1.3 $\times10^5$. To increase the power captured by the light sensitive inductor, we 3D-print an array of 150$\times$150 $\mu$m resin micro lenses on the backside of the sapphire substrate. The shape deviation between design and printed lenses is smaller than 1$\mu$m, and the alignment accuracy of this process is $\delta_x = +5.8 \pm 0.5$ $\mu$m and $\delta_y = +8.3 \pm 3.3$ $\mu$m. We measure a resolving power for 1545-402 nm that is limited to 4.9 by saturation in the KID's phase response. We can model the saturation in the phase response with the evolution of the number of quasiparticles generated by a photon event. An alternative coordinate system that has a linear response raises the resolving power to 5.9 at 402 nm. We verify the measured resolving power with a two-line measurement using a laser source and a monochromator. We discuss several improvements that can be made to the devices on a route towards KID arrays with high resolving powers., Comment: 13 pages, 11 Figures, Journal Paper

Global simulations of electromagnetic turbulence, collisionless tearing modes, and Alfven Eigenmodes in the presence of fast particles are carried out using the gyrokinetic particle-in-cell codes ORB5 (E. Lanti et al, Comp. Phys. Comm., ${\bf 251}$, 107072 (2020)) and EUTERPE (V. Kornilov et al, Phys. Plasmas, ${\bf 11}$, 3196 (2004)) in tokamak and stellarator geometries. Computational feasibility of simulating such complex coupled systems is demonstrated.

World Input-Output (I/O) matrices provide the networks of within- and cross-country economic relations. In the context of I/O analysis, the methodology adopted by national statistical offices in data collection raises the issue of obtaining reliable data in a timely fashion and it makes the reconstruction of (part of) the I/O matrices of particular interest. In this work, we propose a method combining hierarchical clustering and Matrix Completion (MC) with a LASSO-like nuclear norm penalty, to impute missing entries of a partially unknown I/O matrix. Through simulations based on synthetic matrices we study the effectiveness of the proposed method to predict missing values from both previous years data and current data related to countries similar to the one for which current data are obscured. To show the usefulness of our method, an application based on World Input-Output Database (WIOD) tables - which are an example of industry-by-industry I/O tables - is provided. Strong similarities in structure between WIOD and other I/O tables are also found, which make the proposed approach easily generalizable to them.

In the present paper, the evolution of the Alfv\'en modes is studied in a realistic ASDEX Upgrade equilibrium by analysing the results of simulations with the global, electromagnetic, gyrokinetic particle-in-cell code ORB5. The energetic particles are modelled both via the newly implemented isotropic slowing-down and with Maxwellian distribution functions. The comparison of the numerical results shows that modelling the energetic particles with the equivalent Maxwellian rather than with the slowing-down, does not affect the frequency of the driven Alfv\'en mode, while its growth rate appears to be underestimated with a quantitative difference as large as almost 30 %. Additionally the choice of the isotropic slowing-down allows a better description of the nonlinear modification of the dominant Alfv\'en mode frequency, while an equivalent Maxwellian underestimates it. A good comparison with the experimental spectrogram is found.

Hasegawa-Wakatani system, commonly used as a toy model of dissipative drift waves in fusion devices is revisited with considerations of phase and amplitude dynamics of its triadic interactions. It is observed that a single resonant triad can saturate via three way phase locking where the phase differences between dominant modes converge to constant values as individual phases increase in time. This allows the system to have approximately constant amplitude solutions. Non-resonant triads show similar behavior only when one of its legs is a zonal wave number. However when an additional triad, which is a reflection of the original one with respect to the $y$ axis is included, the behavior of the resulting triad pair is shown to be more complex. In particular, it is found that triads involving small radial wave numbers (large scale zonal flows) end up transferring their energy to the subdominant mode which keeps growing exponentially, while those involving larger radial wave numbers (small scale zonal flows) tend to find steady chaotic or limit cycle states (or decay to zero). In order to study the dynamics in a connected network of triads, a network formulation is considered including a pump mode, and a number of zonal and non-zonal subdominant modes as a dynamical system. It was observed that the zonal modes become clearly dominant only when a large number of triads are connected. When the zonal flow becomes dominant as a 'collective mean field', individual interactions between modes become less important, which is consistent with the inhomogeneous wave-kinetic picture. Finally, the results of direct numerical simulation is discussed for the same parameters and various forms of the order parameter are computed. It is observed that nonlinear phase dynamics results in a flattening of the large scale phase velocity as a function of scale in direct numerical simulations.

Energetic particles produced by neutral beams are observed to excite energetic-particle-driven geodesic acoustic modes (EGAMs) in tokamaks. We study the effects of anisotropy of distribution function of the energetic particles on the excitation of such instabilities with ORB5, a gyrokinetic particle-in-cell code. Numerical results are shown for linear electrostatic simulations with ORB5. The growth rate is found to be sensitively dependent on the phase-space shape of the distribution function. The behavior of the instability is qualitatively compared to the theoretical analysis of dispersion relations. Realistic neutral beam energetic particle anisotropic distributions are obtained from the heating solver RABBIT and are introduced into ORB5 as input distribution function. Results show a dependence of the growth rate on the injection angle. A qualitative comparison to experimental measurements is presented and few disagreements between them are found, being the growth rate in the simulations much lower than that from experiments. An explanation for the difference is advanced.

The excitation of toroidicity induced Alfv{\'e}n eigenmodes (TAEs) using prescribed external electromagnetic perturbations (hereafter ``antenna") acting on a confined toroidal plasma as well as its nonlinear couplings to other modes in the system is studied. The antenna is described by an electrostatic potential resembling the target TAE mode structure along with its corresponding parallel electromagnetic potential computed from Ohm's law. Numerically stable long-time linear simulations are achieved by integrating the antenna within the framework of a mixed representation and pullback scheme [A. Mishchenko, et al., Comput. Phys. Commun. \textbf{238} (2019) 194]. By decomposing the plasma electromagnetic potential into symplectic and Hamiltonian parts and using Ohm's law, the destabilizing contribution of the potential gradient parallel to the magnetic field is canceled in the equations of motion. Besides evaluating the frequencies as well as growth/damping rates of excited modes compared to referenced TAEs, we study the interaction of antenna-driven modes with fast particles and indicate their margins of instability. Furthermore, we show first nonlinear simulations in the presence of a TAE-like antenna exciting other TAE modes, as well as Global Alfv\'en Eigenmodes (GAE) having different toroidal wave numbers from that of the antenna.

Electromagnetic turbulence is addressed in tokamak and stellarator plasmas with the global gyrokinetic particle-in-cell codes ORB5 [E. Lanti et al, Comp. Phys. Comm, vol. 251, 107072 (2020)] and EUTERPE [V. Kornilov et al, Phys. Plasmas, vol. 11, 3196 (2004)]. The large-aspect-ratio tokamak, down-scaled ITER, and Wendelstein 7-X geometries are considered. The main goal is to increase the plasma beta, the machine size, the ion-to-electron mass ratio, as well as to include realistic-geometry features in such simulations. The associated numerical requirements and the computational cost for the cases on computer systems with massive GPU deployments are investigated. These are necessary steps to enable electromagnetic turbulence simulations in future reactor plasmas.

This paper presents a study of the interaction between Alfv\'en modes and zonal structures, considering a realistic ASDEX Upgrade equilibrium. The results of gyrokinetic simulations with the global, electromagnetic, particle-in-cell code ORB5 are presented, where the modes are driven unstable by energetic particles with a bump-on-tail equilibrium distribution function, with radial density gradient. Two regimes have been observed: at low energetic particles concentration, the Alfv\'en mode saturates at much higher level in presence of zonal structures; on the other hand at high energetic particles concentration the difference is less pronounced. The former regime is characterized by the zonal structure (identified as an energetic particle driven geodesic acoustic mode), being more unstable than the Alfv\'en mode. In the latter regime the Alfv\'en mode is more unstable than the zonal structure. The theoretical explanation is given in terms of a 3-wave coupling of the energetic particle driven geodesic acoustic mode and Alfv\'en mode, mediated by the curvature-pressure coupling term of the energetic particles.

Pyxel is a novel python tool for end-to-end detection chain simulation i.e. from detector optical effects to readout electronics effects. It is an easy-to-use framework to host and pipeline any detector effect model. It is suited for simulating both Charge-Coupled Devices, CMOS Image Sensors and Mercury Cadmium Telluride hybridized arrays. It is conceived as a collaborative tool to promote reusability, knowledge transfer, and reliability in the instrumentation community. We provide a demonstration of Pyxel's basic principles, describe newly added capabilities, and give examples of more advanced applications.

The nonlinear dynamics of beta-induced Alfv\'en Eigenmodes (BAE) driven by energetic particles (EP) in the presence of ion-temperature-gradient (ITG) turbulence is investigated, by means of selfconsistent global gyrokinetic simulations and analytical theory. A tokamak magnetic equilibrium with large aspect ratio and reversed shear is considered. A previous study of this configuration has shown that the electron species plays an important role in determining the nonlinear saturation level of a BAE in the absence of turbulence [A. Biancalani, et al., J. Plasma Phys. (2020)]. Here, we extend the study to a turbulent plasma. The EPs are found modify the heat fluxes by introducing energy at the large spatial scales, mainly at the toroidal mode number of the dominant BAE and its harmonics. In this regime, BAEs are found to carry a strong electron heat flux. The feed-back of the global relaxation of the temperature profiles induced by the BAE, and on the turbulence dynamics, is also discussed.

The global and electromagnetic gyrokinetic particle-in-cell code ORB5 is employed to investigate the self-consistent interactions between Alfv\'en modes (AM) and ion temperature gradient (ITG) turbulence in a magnetically confined plasma. Here, an axisymmetric magnetic equilibrium with reversed shear and large aspect ratio is considered. An AM with toroidal mode number n=5 is driven unstable by introducing a population of suprathermal ions. Once the AM saturates in the presence of the fully developed turbulence, the ion heat flux is dominated by the AM and its main harmonics. ITG-induced transport is found to also be enhanced in the presence of the unstable AM.

Turbulence in tokamaks generates radially sheared zonal flows. Their oscillatory counterparts, geodesic acoustic modes (GAMs), appear due to the action of the magnetic field curvature. The GAMs can be driven unstable by an anisotropic energetic particle (EP) population leading to the formation of global radial structures, called EGAMs. The EGAMs can redistribute EP energy to the bulk plasma through collisionless wave-particle interaction. In such a way, the EGAMs might contribute to the plasma heating. Thus, investigation of EGAM properties, especially in the velocity space, is necessary for precise understanding of the transport phenomena in tokamak plasmas. In this work, the nonlinear dynamics of EGAMs without considering the mode interaction with the turbulence is investigated with the help of a Mode-Particle-Resonance (MPR) diagnostic implemented in the global gyrokinetic particle-in-cell code ORB5. An ASDEX Upgrade discharge is chosen as a reference case for this investigation due to its rich EP nonlinear dynamics. An experimentally relevant magnetic field configuration, thermal species profiles and an EP density profile are taken for EGAM chirping modelling and its comparison with available empirical data. The same magnetic configuration is used to explore energy transfer by the mode from the energetic particles to the thermal plasma including kinetic electron effects. For a given EGAM level the plasma heating by the mode can be significantly enhanced by varying the EP parameters. Electron dynamics decreases the EGAM saturation amplitude and consequently reduces the plasma heating, even though the mode transfers its energy to thermal ions much more than to electrons.