Search

We introduce MiniMax-01 series, including MiniMax-Text-01 and MiniMax-VL-01, which are comparable to top-tier models while offering superior capabilities in processing longer contexts. The core lies in lightning attention and its efficient scaling. To maximize computational capacity, we integrate it with Mixture of Experts (MoE), creating a model with 32 experts and 456 billion total parameters, of which 45.9 billion are activated for each token. We develop an optimized parallel strategy and highly efficient computation-communication overlap techniques for MoE and lightning attention. This approach enables us to conduct efficient training and inference on models with hundreds of billions of parameters across contexts spanning millions of tokens. The context window of MiniMax-Text-01 can reach up to 1 million tokens during training and extrapolate to 4 million tokens during inference at an affordable cost. Our vision-language model, MiniMax-VL-01 is built through continued training with 512 billion vision-language tokens. Experiments on both standard and in-house benchmarks show that our models match the performance of state-of-the-art models like GPT-4o and Claude-3.5-Sonnet while offering 20-32 times longer context window. We publicly release MiniMax-01 at https://github.com/MiniMax-AI., Comment: A technical report from MiniMax. The authors are listed in alphabetical order. We open-sourced our MiniMax-01 at https://github.com/MiniMax-AI

Medical image segmentation is crucial for accurate clinical diagnoses, yet it faces challenges such as low contrast between lesions and normal tissues, unclear boundaries, and high variability across patients. Deep learning has improved segmentation accuracy and efficiency, but it still relies heavily on expert annotations and struggles with the complexities of medical images. The small size of medical image datasets and the high cost of data acquisition further limit the performance of segmentation networks. Diffusion models, with their iterative denoising process, offer a promising alternative for better detail capture in segmentation. However, they face difficulties in accurately segmenting small targets and maintaining the precision of boundary details. This article discusses the importance of medical image segmentation, the limitations of current deep learning approaches, and the potential of diffusion models to address these challenges., Comment: After a peer review process for a journal submission, we have been told the main conclusions presented in this paper have been proven previously by others. I believe the paper should be withdrawn

Recent advancements in artificial intelligence (AI) have precipitated a paradigm shift in medical imaging, particularly revolutionizing the domain of brain imaging. This paper systematically investigates the integration of deep learning -- a principal branch of AI -- into the semantic segmentation of brain images. Semantic segmentation serves as an indispensable technique for the delineation of discrete anatomical structures and the identification of pathological markers, essential for the diagnosis of complex neurological disorders. Historically, the reliance on manual interpretation by radiologists, while noteworthy for its accuracy, is plagued by inherent subjectivity and inter-observer variability. This limitation becomes more pronounced with the exponential increase in imaging data, which traditional methods struggle to process efficiently and effectively. In response to these challenges, this study introduces the application of adversarial neural networks, a novel AI approach that not only automates but also refines the semantic segmentation process. By leveraging these advanced neural networks, our approach enhances the precision of diagnostic outputs, reducing human error and increasing the throughput of imaging data analysis. The paper provides a detailed discussion on how adversarial neural networks facilitate a more robust, objective, and scalable solution, thereby significantly improving diagnostic accuracies in neurological evaluations. This exploration highlights the transformative impact of AI on medical imaging, setting a new benchmark for future research and clinical practice in neurology.

Neural networks with relatively shallow layers and simple structures may have limited ability in accurately identifying pneumonia. In addition, deep neural networks also have a large demand for computing resources, which may cause convolutional neural networks to be unable to be implemented on terminals. Therefore, this paper will carry out the optimal classification of convolutional neural networks. Firstly, according to the characteristics of pneumonia images, AlexNet and InceptionV3 were selected to obtain better image recognition results. Combining the features of medical images, the forward neural network with deeper and more complex structure is learned. Finally, knowledge extraction technology is used to extract the obtained data into the AlexNet model to achieve the purpose of improving computing efficiency and reducing computing costs. The results showed that the prediction accuracy, specificity, and sensitivity of the trained AlexNet model increased by 4.25 percentage points, 7.85 percentage points, and 2.32 percentage points, respectively. The graphics processing usage has decreased by 51% compared to the InceptionV3 mode.

This research aims to explore the application of deep learning in autonomous driving computer vision technology and its impact on improving system performance. By using advanced technologies such as convolutional neural networks (CNN), multi-task joint learning methods, and deep reinforcement learning, this article analyzes in detail the application of deep learning in image recognition, real-time target tracking and classification, environment perception and decision support, and path planning and navigation. Application process in key areas. Research results show that the proposed system has an accuracy of over 98% in image recognition, target tracking and classification, and also demonstrates efficient performance and practicality in environmental perception and decision support, path planning and navigation. The conclusion points out that deep learning technology can significantly improve the accuracy and real-time response capabilities of autonomous driving systems. Although there are still challenges in environmental perception and decision support, with the advancement of technology, it is expected to achieve wider applications and greater capabilities in the future. potential.

Robust 3D object detection remains a pivotal concern in the domain of autonomous field robotics. Despite notable enhancements in detection accuracy across standard datasets, real-world urban environments, characterized by their unstructured and dynamic nature, frequently precipitate an elevated incidence of false positives, thereby undermining the reliability of existing detection paradigms. In this context, our study introduces an advanced post-processing algorithm that modulates detection thresholds dynamically relative to the distance from the ego object. Traditional perception systems typically utilize a uniform threshold, which often leads to decreased efficacy in detecting distant objects. In contrast, our proposed methodology employs a Neural Network with a self-adaptive thresholding mechanism that significantly attenuates false negatives while concurrently diminishing false positives, particularly in complex urban settings. Empirical results substantiate that our algorithm not only augments the performance of 3D object detection models in diverse urban and adverse weather scenarios but also establishes a new benchmark for adaptive thresholding techniques in field robotics., Comment: This paper has been accepted by the CONF-SEML 2024

Change detection is a fundamental task in computer vision that processes a bi-temporal image pair to differentiate between semantically altered and unaltered regions. Large language models (LLMs) have been utilized in various domains for their exceptional feature extraction capabilities and have shown promise in numerous downstream applications. In this study, we harness the power of a pre-trained LLM, extracting feature maps from extensive datasets, and employ an auxiliary network to detect changes. Unlike existing LLM-based change detection methods that solely focus on deriving high-quality feature maps, our approach emphasizes the manipulation of these feature maps to enhance semantic relevance., Comment: This version is not our full version based on our new progress, related data, and methodology we are dealing with, and based on the rules and the laws, we are adjusting our current version

This study conducts a thorough examination of malware detection using machine learning techniques, focusing on the evaluation of various classification models using the Mal-API-2019 dataset. The aim is to advance cybersecurity capabilities by identifying and mitigating threats more effectively. Both ensemble and non-ensemble machine learning methods, such as Random Forest, XGBoost, K Nearest Neighbor (KNN), and Neural Networks, are explored. Special emphasis is placed on the importance of data pre-processing techniques, particularly TF-IDF representation and Principal Component Analysis, in improving model performance. Results indicate that ensemble methods, particularly Random Forest and XGBoost, exhibit superior accuracy, precision, and recall compared to others, highlighting their effectiveness in malware detection. The paper also discusses limitations and potential future directions, emphasizing the need for continuous adaptation to address the evolving nature of malware. This research contributes to ongoing discussions in cybersecurity and provides practical insights for developing more robust malware detection systems in the digital era.

Research in Cognitive Science suggests that humans understand and represent knowledge of the world through causal relationships. In addition to observations, they can rely on experimenting and counterfactual reasoning -- i.e. referring to an alternative course of events -- to identify causal relations and explain atypical situations. Different instances of control systems, such as smart homes, would benefit from having a similar causal model, as it would help the user understand the logic of the system and better react when needed. However, while data-driven methods achieve high levels of correlation detection, they mainly fall short of finding causal relations, notably being limited to observations only. Notably, they struggle to identify the cause from the effect when detecting a correlation between two variables. This paper introduces a new way to learn causal models from a mixture of experiments on the environment and observational data. The core of our method is the use of selected interventions, especially our learning takes into account the variables where it is impossible to intervene, unlike other approaches. The causal model we obtain is then used to generate Causal Bayesian Networks, which can be later used to perform diagnostic and predictive inference. We use our method on a smart home simulation, a use case where knowing causal relations pave the way towards explainable systems. Our algorithm succeeds in generating a Causal Bayesian Network close to the simulation's ground truth causal interactions, showing encouraging prospects for application in real-life systems., Comment: 10 pages, 13 figures

Deep models, while being extremely flexible and accurate, are surprisingly vulnerable to "small, imperceptible" perturbations known as adversarial attacks. While the majority of existing attacks focus on measuring perturbations under the $\ell_p$ metric, Wasserstein distance, which takes geometry in pixel space into account, has long been known to be a suitable metric for measuring image quality and has recently risen as a compelling alternative to the $\ell_p$ metric in adversarial attacks. However, constructing an effective attack under the Wasserstein metric is computationally much more challenging and calls for better optimization algorithms. We address this gap in two ways: (a) we develop an exact yet efficient projection operator to enable a stronger projected gradient attack; (b) we show that the Frank-Wolfe method equipped with a suitable linear minimization oracle works extremely fast under Wasserstein constraints. Our algorithms not only converge faster but also generate much stronger attacks. For instance, we decrease the accuracy of a residual network on CIFAR-10 to $3.4\%$ within a Wasserstein perturbation ball of radius $0.005$, in contrast to $65.6\%$ using the previous Wasserstein attack based on an \emph{approximate} projection operator. Furthermore, employing our stronger attacks in adversarial training significantly improves the robustness of adversarially trained models., Comment: 30 pages, accepted to ICML 2020

Recently, the conditional SAGE certificate has been proposed as a sufficient condition for signomial positivity over a convex set. In this article, we show that the conditional SAGE certificate is $\textit{complete}$. That is, for any signomial $f(\mathbf{x}) = \sum_{j=1}^{\ell}c_j \exp(\mathbf{A}_j\mathbf{x})$ defined by rational exponents that is positive over a compact convex set $\mathcal{X}$, there is $p \in \mathbb{Z}_+$ and a specific positive definite function $w(\mathbf{x})$ such that $w(\mathbf{x})^p f(\mathbf{x})$ may be verified by the conditional SAGE certificate. The completeness result is analogous to Positivstellensatz results from algebraic geometry, which guarantees representation of positive polynomials with sum of squares polynomials. The result gives rise to a convergent hierarchy of lower bounds for constrained signomial optimization over an $\textit{arbitrary}$ compact convex set that is computable via the conditional SAGE certificate., Comment: 19 pages, preprint

Granular chain packings exhibit a striking emergent strain-stiffening behavior despite the individual looseness of the constitutive chains. Using indentation experiments on such assemblies, we measure an exponential increase in the collective resistance force $F$ with the indentation depth $z$, and with the square root of the number $\mathcal{N}$ of beads per chain. These two observations are respectively reminiscent of the self-amplification of friction in a capstan or in interleaved books, as well as the physics of polymers. The experimental data are well captured by a novel model based on these two ingredients. Specifically, the resistance force is found to vary according to the universal relation: $\log F \sim \mu \sqrt{\mathcal{N}} \Phi^{11/8}z/ b $, where $\mu$ is the friction coefficient between two elementary beads, $b$ is their size, and $\Phi$ is the volume fraction of chain beads when semi-diluted in a surrounding medium of unconnected beads. Our study suggests that theories normally confined to the realm of polymer physics at a molecular level can be used to explain phenomena at a macroscopic level. This class of systems enables the study of friction in complex assemblies, with practical implications for the design of new materials, the textile industry, and biology.

Multivariate classification methods using explanatory and predictive models are necessary for characterizing subgroups of patients according to their risk profiles. Popular methods include logistic regression and classification trees with performances that vary according to the nature and the characteristics of the dataset. In the context of imported malaria, we aimed at classifying severity criteria based on a heterogeneous patient population. We investigated these approaches by implementing two different strategies: L1 logistic regression (L1LR) that models a single global solution and classification trees that model multiple local solutions corresponding to discriminant subregions of the feature space. For each strategy, we built a standard model, and a sparser version of it. As an alternative to pruning, we explore a promising approach that first constrains the tree model with an L1LR-based feature selection, an approach we called L1LR-Tree. The objective is to decrease its vulnerability to small data variations by removing variables corresponding to unstable local phenomena. Our study is twofold: i) from a methodological perspective comparing the performances and the stability of the three previous methods, i.e L1LR, classification trees and L1LR-Tree, for the classification of severe forms of imported malaria, and ii) from an applied perspective improving the actual classification of severe forms of imported malaria by identifying more personalized profiles predictive of several clinical criteria based on variables dismissed for the clinical definition of the disease. The main methodological results show that the combined method L1LR-Tree builds sparse and stable models that significantly predicts the different severity criteria and outperforms all the other methods in terms of accuracy., Comment: 18 pages, 10 figures, ICLR, computational science - Learning, Imported Malaria, L1 logistic regression, Decision tree

Structural, elastic and thermal properties of cementite (Fe$_3$C) were studied using a Modified Embedded Atom Method (MEAM) potential for iron-carbon (Fe-C) alloys. Previously developed Fe and C single element potentials were used to develop an Fe-C alloy MEAM potential, using a statistically-based optimization scheme to reproduce structural and elastic properties of cementite, the interstitial energies of C in bcc Fe as well as heat of formation of Fe-C alloys in L$_{12}$ and B$_1$ structures. The stability of cementite was investigated by molecular dynamics simulations at high temperatures. The nine single crystal elastic constants for cementite were obtained by computing total energies for strained cells. Polycrystalline elastic moduli for cementite were calculated from the single crystal elastic constants of cementite. The formation energies of (001), (010), and (100) surfaces of cementite were also calculated. The melting temperature and the variation of specific heat and volume with respect to temperature were investigated by performing a two-phase (solid/liquid) molecular dynamics simulation of cementite. The predictions of the potential are in good agreement with first-principles calculations and experiments., Comment: 12 pages, 9 figures

A set of Modified Embedded Atom Method (MEAM) potentials for the interactions between Al, Si, Mg, Cu, and Fe was developed from a combination of each element's MEAM potential in order to study metal alloying. Previously published MEAM parameters of single elements have been improved for better agreement to the generalized stacking fault energy (GSFE) curves when compared with ab-initio generated GSFE curves. The MEAM parameters for element pairs were constructed based on the structural and elastic properties of element pairs in the NaCl reference structure garnered from ab-initio calculations, with adjustment to reproduce the ab-initio heat of formation of the most stable binary compounds. The new MEAM potentials were validated by comparing the formation energies of defects, equilibrium volumes, elastic moduli, and heat of formation for several binary compounds with ab-initio simulations and experiments. Single elements in their ground state crystal structure were subjected to heating to test the potentials at elevated temperatures. An Al potential was modified to avoid formation of an unphysical solid structure at high temperatures. The thermal expansion coefficient of a compound with the composition of AA 6061 alloy was evaluated and compared with experimental values. MEAM potential tests performed in this work, utilizing the universal Atomistic Simulation Environment (ASE), are distributed to facilitate reproducibility of the results., Comment: 17 pages, 5 figures, for PRB

We report a first-principles calculation that models the effect of iron (Fe) atoms on the adsorption of a tungsten (W) atom on W(100) surfaces. The adsorption of a W atom on a clean W(100) surface is compared with that of a W atom on a W(100) surface covered with a monolayer of Fe atoms. The total energy of the system is computed as the function of the height of the W adatom. Our result shows that the W atom first adsorbs on top of the Fe monolayer. Then the W atom can replace one of the Fe atoms through a path with a moderate energy barrier and reduce its energy further. This intermediate site makes the adsorption (and desorption) of W atoms a two-step process in the presence of Fe atoms and lowers the overall adsorption energy by nearly 2.4 eV. The Fe atoms also provide a surface for W atoms to adsorb facilitating the diffusion of W atoms. The combination of these two effects result in a much more efficient desorption and diffusion of W atoms in the presence of Fe atoms. Our result provides a fundamental mechanism that can explain the activated sintering of tungsten by Fe atoms., Comment: 9 pages, 2 figures

We have developed a multi-objective optimization (MOO) procedure to construct modified-embedded-atom-method (MEAM) potentials with minimal manual fitting. This procedure has been applied successfully to develop a new MEAM potential for magnesium. The MOO procedure is designed to optimally reproduce multiple target values that consist of important materials properties obtained from experiments and first-principles calculations based on density-functional theory (DFT). The optimized target quantities include elastic constants, cohesive energies, surface energies, vacancy formation energies, and the forces on atoms in a variety of structures. The accuracy of the new potential is assessed by computing several material properties of Mg and comparing them with those obtained from other potentials previously published. We found that the present MEAM potential yields a significantly better overall agreement with DFT calculations and experiments., Comment: 8 pages, 2 figures

We use density-functional theory to study the structure of AlSb(001) and GaSb(001) surfaces. Based on a variety of reconstruction models, we construct surface stability diagrams for AlSb and GaSb under different growth conditions. For AlSb(001), the predictions are in excellent agreement with experimentally observed reconstructions. For GaSb(001), we show that previously proposed model accounts for the experimentally observed reconstructions under Ga-rich growth conditions, but fails to explain the experimental observations under Sb-rich conditions. We propose a new model that has a substantially lower surface energy than all (nx5)-like reconstructions proposed previously and that, in addition, leads to a simulated STM image in better agreement with experiment than existing models. However, this new model has higher surface energy than some of (4x3)-like reconstructions, models with periodicity that has not been observed. Hence we conclude that the experimentally observed (1x5) and (2x5) structures on GaSb(001) are kinetically limited rather than at the ground state., Comment: 6 pages, 6 figures

We report a molecular dynamics simulation of melting of tungsten (W) nanoparticles. The modified embedded atom method (MEAM) interatomic potentials are used to describe the interaction between tungsten atoms. The melting temperature of unsupported tungsten nanoparticles of different sizes are found to decrease as the size of the particles decreases. The melting temperature obtained in the present study is approximately a decreasing function of inverse radius, in a good agreement with the predictions of thermodynamic models. We also observed that the melting of a W nanoparticle is preceded by the premelting of its outer skin at a temperature lower than its melting temperature., Comment: 7 pages, 4 figures

We developed new modified embedded-atom method (MEAM) interatomic potentials for the Mg-Al alloy system using a first-principles method based on density functional theory (DFT). The materials parameters, such as the cohesive energy, equilibrium atomic volume, and bulk modulus, were used to determine the MEAM parameters. Face-centered cubic, hexagonal close packed, and cubic rock salt structures were used as the reference structures for Al, Mg, and MgAl, respectively. The applicability of the new MEAM potentials to atomistic simulations for investigating Mg-Al alloys was demonstrated by performing simulations on Mg and Al atoms in a variety of geometries. The new MEAM potentials were used to calculate the adsorption energies of Al and Mg atoms on Al (111) and Mg (0001) surfaces. The formation energies and geometries of various point defects, such as vacancies, interstitial defects and substitutional defects, were also calculated. We found that the new MEAM potentials give a better overall agreement with DFT calculations and experiments when compared against the previously published MEAM potentials., Comment: Fixed a reference

The validation of NASA Global Precipitation Mission (GPM) satellite precipitation products is important for their credibility and utility within the larger community. This document defines GPM ground validation scientific objectives and several programmatic components for meeting those objectives. Multi-year, multi-sensor ground-based observation programs in a few locations are proposed to generate local observation products and global error covariance products. Focused measurement programs utilizing aircraft, ships, and ground-based measurements would fill in geographic and scientific gaps not addressed by the multi-year observing programs., Comment: 22 pages, 1 figure. 3 April 2002 version. Submitted for inclusion in GPM Draft Science Implementation Plan