Publications

A new BP-LSTM hybrid wind power forecasting model based on improved SFLA optimization
Houjian Zhan
Jiaqiu Hu
Le Qi
Xinran Liu
Xianliang Cheng
Yan Chen
Zhenhua Xiong
To improve the accuracy of short-term wind power prediction, this paper proposes a short-term wind power hybrid forecasting model, ISFLA-BP-… (see more)LSTM, based on an improved Shuffled Frog Leaping Algorithm (ISFLA) optimized Back Propagation network (BP) and Long Short-term Memory network (LSTM). This method introduces an adaptive differential evolution operator to optimize local search operations, enhancing the ISFLA optimization algorithm's ability to escape "local optimal solutions" and speeding up convergence. Finally, the improved ISFLA algorithm is used to optimize the weights and biases of BP and the hyperparameters of LSTM separately, establishing short-term wind power prediction submodels. The final short-term wind power prediction values are obtained by integrating the power outputs of ISFLA-BP and ISFLA-LSTM. This paper validates the performance of ISFLA and ISFLA-BP-LSTM on five commonly used test functions and wind energy data collected from the Xiyi Mountain Wind Farm in the Guangxi Zhuang Autonomous Region of China. The experimental results indicate that ISFLA in this paper outperforms the original SFLA in terms of performance, and the ISFLA-BP-LSTM hybrid model exhibits higher prediction accuracy compared to the initial model.
Probable Detection of a Cooler Gas Component in the Perseus Cluster with XRISM
Julian Meunier
Brian R. McNamara
A. Simionescu
François Mernier
Irina Zhuravleva
Congyao Zhang
Annie Heinrich
Frederick Scott Porter
John A. ZuHone
Elena Bellomi
Ian Drury
Megan E. Eckart
Ryuichi Fujimoto
Yutaka Fujita
Liyi Gu
Isamu Hatsukade
Yuto Ichinohe
Yoshiaki Kanemaru … (see 19 more)
Takao Kitaguchi
Shunji Kitamoto
Shōgo Kobayashi
Takayoshi Kohmura
Hironori Matsumoto
Kyoko Matsushita
K. Migkas
Ikuyuki Mitsuishi
Koji Mori
Hiroshi Nakajima
Hirofumi Noda
Anna Ogorzałek
Naomi Ota
Lior Shefler
Nhut Truong
Ayşegül Tümer
Nagomi Uchida
Yuusuke Uchida
Shutaro Ueda
Abstract We present an analysis of the temperature structure of the Perseus cluster atmosphere using X-Ray Imaging and Spectroscopy Mission … (see more)(XRISM) Resolve observations. The average temperature rises from 3.3 keV near the nucleus of NGC 1275 to 8 keV at 10′ (210 kpc), which is consistent with Chandra and XMM measurements. The velocity and velocity dispersion profiles are broadly consistent with those in XRISM Collaboration et al. While the gas at altitudes beyond ∼60 kpc can be modeled as a single temperature plasma, we find evidence for more than one gas phase in the inner ∼60 kpc. The hotter gas component, traced primarily by the Fe He α line, has a velocity dispersion of ≲140 km s −1 . We detect a cooler, ∼1.87–2.43 keV, gas component with a velocity dispersion of ∼300–400 km s −1 , a bulk velocity of ∼21–213 km s −1 with respect to the central galaxy, and a mass of 8 × 10 9 –13 × 10 10 M ⊙ . These ranges reflect large systematic uncertainties associated with modeling spatial-spectral mixing and the bright central point source. Potential low energy gain variations may add additional uncertainties. The cooler component is identified by broad wings in prominent emission lines, most notably S Ly α and Fe He α . This cooler component’s Mach number ∼0.73–0.96 and nonthermal pressure fraction of ∼22.9%–33.7% are much higher than found for the hotter gas. The cooler gas may be associated with merging halos along the line of sight, which formed the cool, sloshing spiral and/or cooling gas being disturbed by the radio jets and lobes.
A Profit-Driven Simulation framework for real-time evaluation of toxicity detection models in social media
Arezo Bodaghi
Benjamin C. M. Fung
Jonathan Shahen
Ketra A. Schmitt
SnowGalileo: A Pre-trained Earth Observation Transformer for Daily, 100 m Fractional Snow Cover Mapping
Donovan J. M. Allum
Sebastian Roessler
Samip Shrestha
Zhibang Lv
J Truckenbrodt
Gabriele Schwaizer
Thomas Nägler
John W. Pomeroy
Christopher B. Marsh
Benoît Montpetit
Tobias Jonas
A.J. Dietz
Celia A. Baumhoer
Abstract. Mountain snow is an important component of the cryosphere that directly affects downstream livelihoods. Accurate monitoring of mou… (see more)ntain snow is crucial, yet remains challenging due to complex topography and frequent cloud cover. Although combining data from multiple Earth Observation (EO) satellites can improve spatial and temporal coverage, extracting Fractional Snow Cover (FSC) from sensors with different spatial and temporal resolutions remains difficult. AI-based Earth foundation models can process diverse sensor inputs and have been successfully applied to various remote sensing tasks; however, they often lack snow-specific design considerations. This paper presents SnowGalileo, a pre-trained transformer model designed to integrate EO data to map snow-covered areas. We evaluated SnowGalileo's potential for generating daily, gap-free FSC maps at 100m resolution in mountainous regions. SnowGalileo combines one week of satellite time-series data from multiple sensors, including Sentinel-1, Sentinel-2, Landsat, Sentinel-3, MODIS, and VIIRS, with topographic, land cover, and meteorological information to predict FSC for a given day. The model was pre-trained using masked autoencoding and fine-tuned using labeled data from various mountain ranges across the Northern Hemisphere. In addition to clear-sky conditions, SnowGalileo is evaluated under a wider variety of conditions than was previously possible, including cloud cover and lack of high-resolution (~10–30 m) satellite imagery. For the Canadian Rockies and the Swiss Alps, respectively, SnowGalileo achieves RMSEs of 0.099 and 0.124 on clear days, 0.144 and 0.209 on cloudy days, 0.185 and 0.262 on days without high-resolution imagery, and 0.200 and 0.299 on cloudy days without high-resolution imagery. SnowGalileo also consistently outperforms random forests, support vector regressors, and multi-layer perceptrons. While the current product is a proof of concept that has undergone limited validation across geographic regions, operates on 1km × 1km tiles rather than full maps, and has restricted capabilities in challenging conditions, this approach could ultimately enable the continuous, gap-free operational generation of FSC time series for mountain regions worldwide.
Evaluating the Performance of Traditional Pharmacoepidemiologic and Machine Learning Models to Predict Pregnancies at Risk of Major Congenital Malformations
Gabra Nohmie
Marc J. Lanovaz
Odile Sheehy
Cristina Longo
Robert W. Platt
Christine Damase‐Michel
Anick Bérard
BACKGROUND: With approximately 50% of pregnancies being unplanned, there is an unintended exposure to potential feto-toxic drugs that may ca… (see more)use major congenital malformations (MCM). This study aims to compare the predictive performance between traditional pharmacoepidemiologic (PE) and machine learning (ML) models. METHODS: We conducted a cohort study within the Quebec Pregnancy Cohort, including all pregnancies covered by Quebec's prescription drug insurance program and their children from 01/1998 to 12/2015. Medication exposures, comorbidities, and women's characteristics 12 months before pregnancy and during the first trimester were considered. Robust Poisson models were used to obtain adjusted risk ratios (aRR) and 95% confidence intervals (CI) of predictors. Logistic regression, robust Poisson, K-Nearest Neighbors, Random Forest, XGBoost, Naïve bayes, Multilayer Perceptron, and Support Vector Machine were developed to predict pregnancies at risk of MCM. Sensitivity, specificity, PPV, NPV, accuracy, ROC-AUCs, PR-AUCs, and F1-score were used to evaluate the performance of predictive models. RESULTS: We analyzed 213,744 pregnancies, finding a 9.7% prevalence of MCM. Logistic regression had the highest discriminative power across models at predicting MCM, with a ROC-AUC of 53.3% and the highest sensitivity (41.5%) and F1-score (46.6%). KNN had the highest specificity (97.9%) but the lowest sensitivity (2.3%). Robust Poisson performed similarly to logistic regression, with the highest accuracy (52.3%). Robust Poisson performed slightly better than logistic regression at classifying organ-specific malformations. All models showed poor overall predictive performance. Results were robust across sensitivity analyses. CONCLUSION: There is insufficient evidence for the superiority of ML over traditional pharmacoepidemiologic modeling in predicting MCM.
FrogNano: Training a 4B Coding Agent via Online Task Synthesis
Minseon Kim
Zhengyan Shi
Christopher Cui
Roger Creus Castanyer
Isadora White
Jonathan Light
Jeonghye Kim
Matheus Pereira
Darya Moldavskaya
Chinmay Singh
Fabio Vera
Baolin Peng
Xingdi Yuan
We present FrogNano, a 4B coding agent designed to tackle software engineering (SWE) tasks efficiently and effectively, even under resource-… (see more)constrained environments. It is post-trained exclusively via RL on around 1,500 SWE environments with synthetic tasks. A key ingredient for improving performance is an online task synthesis pipeline that creates tasks calibrated to the frontier of learnability for the current checkpoint. This report provides evidence that competitive small coding agents can be trained with synthetic tasks alone, without traditional distillation from larger models, and that generating tasks at the learnability frontier of the current agent is important. We report details on the training methodology, evaluations across diverse environments, and in-depth analyses, serving as a foundation for our ongoing exploration of lightweight yet capable coding agents that can run on minimal hardware.
Observation-driven simulations of strong lensing galaxy clusters
L. Leuzzi
Massimo Meneghetti
A. Adam
Lauro Moscardini
C. Giocoli
P. Bergamini
Matteo Maturi
Amata Mercurio
A. Moretti
Andrés Alejandro Plazas Malagón
P. Rosati
Galaxy clusters are the most powerful strong lenses: they greatly magnify the flux of distant and faint sources. Strong lensing also allows … (see more)for the reconstruction of their mass distribution with-1% level accuracy and enables investigating cosmological parameters. The number of known systems of this type is bound to increase in the next years thanks to wide imaging surveys. Using simulations in this context is crucial to validate the analysis methods before they are applied to real data, and to train machine learning algorithms that can handle large volumes of images. In this work, we present a simulated set of one hundred images of galaxy clusters that we have produced with a novel code for simulating cluster-scale strong lenses. One of the main novelties of our approach, distinguishing it from other existing codes, is the use of empirical relations, derived from state-of-the-art observations, for modelling the characteristics, such as morphology, color, and spatial distribution of the cluster member population. This allows us to reliably reproduce the complexity of real observations. The simulations are partly carried out with the latest version of SkyLens, a code that creates mock observations of strong lensing events in different systems and observational setups. The main improvements we introduce are the use of the message passage interface (MPI) paradigm, a standard for parallel programming that leverages the use of several processors to perform some given task, and the implementation of score-based diffusion models to augment the images of the background sources. Together, they lead to more efficient and realistic image simulations. We also present the validation of the code and simulations by comparing the properties of the mock clusters to those of real ones. [Abridged]
ATAD2 is a novel regulator of myogenesis and autophagy in skeletal muscles
Sami Sedraoui
Alaa Moamer
Tomer Jordi Chaffer
Jean-Philippe Leduc-Gaudet
Dominique Mayaki
Hanfen Shi
Ryann Lang
Minna Woo
Yumin Zheng
Marco Sandri
Gilles Gouspillou
Sabah N.A. Hussain
The nuclear protein ATAD2 (ATPase family AAA domain containing 2) is a known positive regulator of cell proliferation in various cancer type… (see more)s. The expression and functional roles of ATAD2 in skeletal muscle cells are unknown. In this study, we used transient and stable knockdown approaches using siRNA and shRNA oligos to evaluate how ATAD2 regulates proliferation, migration, differentiation, and autophagy in C2C12 myoblasts. Atad2 knockdown (KD) significantly increased myoblast proliferation rate, S-phase entry, overall cell viability, and early differentiation into myotubes. However, Atad2 KD also elicited myotube atrophy and upregulation of ubiquitin E3 ligases Atrogin-1 and MuRF1. Basal autophagy was inhibited in Atad2 KD cells as a result of downregulation of autophagy-related genes ( Lc3b , Gabarapl1 , Atg5 , and Atg7 ). Immunoblotting and immunostaining revealed significant decrease in LC3B protein levels and the number of LC3B punctae per cell with Atad2 KD, respectively. LAMP1 staining confirmed the presence of enlarged lysosomes in Atad2 KD cells. Additionally, genes involved in lysosome function and integrity including Rab7 , Rab29 , Rab32 , Nrbf2 , and Cathepsin L were downregulated in Atad2 KD cells. Collectively, these results indicate that ATAD2 is a critical regulator of muscle cell proliferation, differentiation, autophagy and lysosomal integrity.
From Visual Widgets to UI Code: Efficient Tool-Grounded Generation
Houston H. Zhang
Linfeng Ye
Yuanhao Yu
Xinxin Zuo
Zhixiang Chi
Existing screenshot-to-code systems face a trade-off between flexibility and controllability. Direct multimodal generation can hallucinate v… (see more)isible details, whereas structured pipelines reduce such errors through component-wise decomposition, predefined templates, and customized intermediate representations. These structures, however, introduce additional generative orchestration and restrict outputs to designs covered by the representation. We investigate whether selective tool grounding can improve the fidelity--efficiency trade-off of direct widget-to-code generation. We introduce \textbf{WidgetGen}, a lightweight tool-grounded framework that extracts observable text and color evidence, performs high-level layout and optional chart reasoning, and directly generates executable JavaScript XML (\emph{JSX}). This design reduces reliance on component-wise generation while avoiding a fixed UI schema. Across six multimodal models and \(1{,}000\) held-out widgets, WidgetGen outperforms direct prompting and the structured Widget2Code pipeline on most visual reconstruction metrics, with consistent gains in area, legibility, and style. Finally, reconstruction-derived image-code pairs improve six Qwen-family open-weight models across every reported metric through supervised fine-tuning. These results establish WidgetGen as a strong lightweight baseline and show that selective evidence grounding offers an effective alternative to extensive representation constraints.
Morphology-Conditioned World Model for Cross-Embodiment Quadrupedal Locomotion
World models promise a paradigm shift in robotics, where an agent learns the physics of its environment once and then acquires behaviors eff… (see more)iciently. Yet the learned dynamics models at their core are typically morphology locked. In legged locomotion, a dynamics model trained on an ANYmal-D quadruped fails on a Unitree Go1 because it overfits to one robot's embodiment rather than capturing the locomotion dynamics shared across robots, so even a small change in actuator dynamics or limb length forces retraining from scratch. However, if we formalize a robot's unique physical traits into a morphology specification, a controller for a family of robots can utilize this blueprint in two ways. It can feed the specification to a model-free policy, or it can feed the specification to a learned dynamics model and extract the policy in imagination. We argue for the second route and introduce the Quadrupedal World Model (QWM), which conditions a single generative dynamics model on scale-invariant physical features and trains policies entirely inside it, through a physical morphology encoder, an adaptive reward normalizer, and morphology conditioning in the latent dynamics. Holding the morphology information identical, a model-free policy matches QWM on the training cohort but degrades on unseen morphologies, while QWM transfers zero-shot with no fine-tuning, adaptation, or warm-up in such cases. To our knowledge, this is the first world model to demonstrate zero-shot cross-embodiment transfer within the quadrupedal family.
Scaling Self-Play for End-to-End Driving
Daphne Cornelisse
Felix Heide
Eugene Vinitsky
Christopher Pal
End-to-end autonomous driving models are typically trained on offline human-demonstration datasets that provide limited state coverage and o… (see more)ften no closed-loop feedback, making them prone to compounding errors when deployed in closed-loop and brittle to long-tail agent interactions. To overcome these limitations, we propose an alternative strategy for training end-to-end driving models: large-scale self-play directly from pixels in simulation. While prior self-play approaches have shown promising transfer to real-world driving, they typically assume vectorized Bird's-Eye-View (BEV) observations that are incompatible with end-to-end policies operating directly on sensor observations. To this end, we introduce Gigapixel, a high-throughput batched driving simulator with perspective rendering, enabling scalable self-play directly from pixel observations. Rather than targeting compute-costly photorealistic sensor simulation, Gigapixel renders a simplified bounding-box world that preserves essential scene structure while achieving throughput at 50k agent steps per second. Since direct pixel-space self-play RL is prohibitively sample-inefficient at end-to-end model scale, we propose self-play DAgger training: we train pixel-based policies in self-play via on-policy distillation from a privileged RL teacher. To bridge the sim-to-real gap, we subsequently transfer the self-play trained policies to real-world sensor data through lightweight perception adaptation. Policies trained in Gigapixel and adapted to real-world sensor data achieve competitive performance on the HUGSIM and NAVSIM-v2 benchmarks without human trajectory supervision. Moreover, scaling self-play training yields proportional gains in policy performance, establishing self-play as a practical and scalable strategy for training end-to-end models.
From Complex Behavior to Intelligent Human-Centered AI: The 11th ABAW Workshop & Competition
Dimitrios Kollias
Stefanos Zafeiriou
Irene Kotsia
Alessandro Lameiras Koerich
Simon L Bacon
Oya Celiktutan
Soufiane Belharbi
Muhammad Osama Zeeshan
Muhammad Haseeb Aslam
Chunchang Shao
Guanyu Hu
The 11th Affective & Behavior Analysis in-the-Wild (ABAW) Workshop and Competition, held in conjunction with ECCV 2026, continues to advance… (see more) research on the modelling, analysis and understanding of human affect and behavior in unconstrained real-world environments. The workshop maintains its dual structure, comprising both a competition and a paper track. The ABAW Competition features two challenging benchmarks targeting key problems in affective and behavioral understanding: Multi-Task Learning (MTL), requiring the joint estimation of valence-arousal, facial expressions and action units from a static version of the Aff-Wild2 database, and Ambivalence/Hesitancy (AH) Video Recognition, focusing on the recognition of complex behavioral states from multimodal videos using the BAH dataset. In parallel, the paper track presents recent advances spanning multimodal AI and deep learning, affective computing, behavior understanding, datasets and benchmarks, human-centered AI, human-robot interaction and social robotics, robustness, explainability and real-world applications. Overall, the 11th ABAW Workshop and Competition continues to serve as a leading forum for benchmarking, collaboration and innovation, fostering the development of the next generation of multimodal, trustworthy and human-centered AI systems.