arXiv subjects

Masato Taki

Publications and source records attributed to Masato Taki.

At least 19 recordsLinked to original sources

arXiv2026

Phases in a class of associative memories via hidden neurons

Associative memory in the Hopfield network is attractor dynamics in a disordered many-body system, and higher-order and exponential extensions turn its retrieval update into softmax attention. The polynomial and exponential regimes have been analyzed by different methods, with no common architecture in which to ask what fixes the storage scale. In this paper we study the bipartite architecture of Krotov and Hopfield, which we call the class $H$, whose model is fixed by a Lagrangian for each layer, taking the hidden neurons as the order parameter of retrieval. At polynomial load the replica method yields the replica-symmetric phase diagrams and closed-form capacities, and the crosstalk moment is common to Ising and spherical visible neurons, so their differences come from the visible entropy. With a softmax hidden layer the load is exponential, and a copy representation maps the thermodynamics onto random-energy-model counting, with paramagnetic, condensed, and frozen phases. Heating destabilizes retrieval by quantized reassignments of attention, and typical Gaussian patterns remain metastable at every load. The regimes differ in their crosstalk statistics, central-limit at polynomial load and large-deviation at exponential load, and the class $H$ splits retrieval into two roles, the visible Lagrangian fixing stability and the hidden one the storage scale, two axes that may also guide the design of new Lagrangians.

cs.LG↗

arXiv2025

On the Role of Hidden States of Modern Hopfield Network in Transformer

Associative memory models based on Hopfield networks and self-attention based on key-value mechanisms have been popular approaches in the study of memory mechanisms in deep learning. It has been pointed out that the state update rule of the modern Hopfield network (MHN) in the adiabatic approximation is in agreement with the self-attention layer of Transformer. In this paper, we go beyond this approximation and investigate the relationship between MHN and self-attention. Our results show that the correspondence between Hopfield networks and Transformers can be established in a more generalized form by adding a new variable, the hidden state derived from the MHN, to self-attention. This new attention mechanism, modern Hopfield attention (MHA), allows the inheritance of attention scores from the input layer of the Transformer to the output layer, which greatly improves the nature of attention weights. In particular, we show both theoretically and empirically that MHA hidden states significantly improve serious problem of deep Transformers known as rank collapse and token uniformity. We also confirm that MHA can systematically improve accuracy without adding training parameters to the Vision Transformer or GPT. Our results provide a new case in which Hopfield networks can be a useful perspective for improving the Transformer architecture.

cs.LG↗

arXiv2025

GeoProg3D: Compositional Visual Reasoning for City-Scale 3D Language Fields

The advancement of 3D language fields has enabled intuitive interactions with 3D scenes via natural language. However, existing approaches are typically limited to small-scale environments, lacking the scalability and compositional reasoning capabilities necessary for large, complex urban settings. To overcome these limitations, we propose GeoProg3D, a visual programming framework that enables natural language-driven interactions with city-scale high-fidelity 3D scenes. GeoProg3D consists of two key components: (i) a Geography-aware City-scale 3D Language Field (GCLF) that leverages a memory-efficient hierarchical 3D model to handle large-scale data, integrated with geographic information for efficiently filtering vast urban spaces using directional cues, distance measurements, elevation data, and landmark references; and (ii) Geographical Vision APIs (GV-APIs), specialized geographic vision tools such as area segmentation and object detection. Our framework employs large language models (LLMs) as reasoning engines to dynamically combine GV-APIs and operate GCLF, effectively supporting diverse geographic vision tasks. To assess performance in city-scale reasoning, we introduce GeoEval3D, a comprehensive benchmark dataset containing 952 query-answer pairs across five challenging tasks: grounding, spatial reasoning, comparison, counting, and measurement. Experiments demonstrate that GeoProg3D significantly outperforms existing 3D language fields and vision-language models across multiple tasks. To our knowledge, GeoProg3D is the first framework enabling compositional geographic reasoning in high-fidelity city-scale 3D environments via natural language. The code is available at https://snskysk.github.io/GeoProg3D/.

cs.CV↗

arXiv2025

Artificial intelligence pioneers the double-strangeness factory

Artificial intelligence (AI) is transforming not only our daily experiences but also the technological development landscape and scientific research. In this study, we pioneered the application of AI in double-strangeness hypernuclear studies. These studies which investigate quantum systems with strangeness via hyperon interactions provide insights into fundamental baryon-baryon interactions and contribute to our understanding of the nuclear force and composition of neutron star cores. Specifically, we report the observation of a double hypernucleus in nuclear emulsion achieved via innovative integration of machine learning techniques. The proposed methodology leverages generative AI and Monte Carlo simulations to produce training datasets combined with object detection AI for effective event identification. Based on the kinematic analysis and charge identification, the observed event was uniquely identified as the production and decay of resulting from {\Xi}- capture by 14N in the nuclear emulsion. Assuming capture in the atomic 3D state, the binding energy of the two {\Lambda} hyperons in 13B{\Lambda}{\Lambda}, B{\Lambda}{\Lambda}, was determined as 25.57 +- 1.18(stat.) +- 0.07(syst.) MeV. The {\Lambda}{\Lambda} interaction energy obtained was 2.83 +- 1.18(stat.) +- 0.14(syst.) MeV. This study marks a new era in double-strangeness research.

nucl-ex↗

arXiv2025

Binding energy of $^{3}_{\Lambda}\rm{H}$ and $^{4}_{\Lambda}\rm{H}$ via image analyses of nuclear emulsions using deep-learning

Subatomic systems are pivotal for understanding fundamental baryonic interactions, as they provide direct access to quark-level degrees of freedom. In particular, introducing a strange quark adds "strangeness" as a new dimension, offering a powerful tool for exploring nuclear forces. The hypertriton, the lightest three-body hypernuclear system, provides an ideal testing ground for investigating baryonic interactions and quark behavior involving up, down, and strange quarks. However, experimental measurements of its lifetime and binding energy, key indicators of baryonic interactions, show significant deviations in results obtained from energetic collisions of heavy-ion beams. Identifying alternative pathways for precisely measuring the hypertriton's binding energy and lifetime is thus crucial for advancing experimental and theoretical nuclear physics. Here, we present an experimental study on the binding energies of $^3_{\Lambda}\mathrm{H}$ (hypertriton) and $^4_{\Lambda}\mathrm{H}$, performed through the analysis of photographic nuclear emulsions using modern techniques. By incorporating deep-learning methods, we uncovered systematic uncertainties in conventional nuclear emulsion analyses and established a refined calibration protocol for determining binding energies accurately. Our results are independent of those obtained from heavy-ion collision experiments, offering a complementary measurement and opening new avenues for investigating few-body hypernuclei interactions.

nucl-ex↗

arXiv2024

A novel machine learning method to detect double-$\Lambda$ hypernuclear events in nuclear emulsions

A novel method was developed to detect double-$\Lambda$ hypernuclear events in nuclear emulsions using machine learning techniques. The object detection model, the Mask R-CNN, was trained using images generated by Monte Carlo simulations, image processing, and image-style transformation based on generative adversarial networks. Despite being exclusively trained on $\prescript{6\ }{\Lambda\Lambda}{\rm{He}}$ events, the model achieved a detection efficiency of 93.8$\%$ for $\prescript{6\ }{\Lambda\Lambda}{\rm{He}}$ and 82.0$\%$ for $\prescript{5\ }{\Lambda\Lambda}{\rm{H}}$ events in the produced images. In addition, the model demonstrated its ability to detect the $\prescript{6\ }{\Lambda\Lambda}{\rm{He}}$ event named the Nagara event, which is the only uniquely identified double-$\Lambda$ hypernuclear event reported to date. It also exhibited a proper segmentation of the event topology. Furthermore, after analyzing 0.2$\%$ of the entire emulsion data from the J-PARC E07 experiment utilizing the developed approach, six new candidates for double-$\Lambda$ hypernuclear events were detected, suggesting that more than 2000 double-strangeness hypernuclear events were recorded in the entire dataset. This method is sufficiently effective for mining more latent double-$\Lambda$ hypernuclear events recorded in nuclear emulsion sheets by reducing the time required for manual visual inspection by a factor of five hundred.

hep-ex↗

arXiv2024

Hierarchical Associative Memory, Parallelized MLP-Mixer, and Symmetry Breaking

Transformers have established themselves as the leading neural network model in natural language processing and are increasingly foundational in various domains. In vision, the MLP-Mixer model has demonstrated competitive performance, suggesting that attention mechanisms might not be indispensable. Inspired by this, recent research has explored replacing attention modules with other mechanisms, including those described by MetaFormers. However, the theoretical framework for these models remains underdeveloped. This paper proposes a novel perspective by integrating Krotov's hierarchical associative memory with MetaFormers, enabling a comprehensive representation of the entire Transformer block, encompassing token-/channel-mixing modules, layer normalization, and skip connections, as a single Hopfield network. This approach yields a parallelized MLP-Mixer derived from a three-layer Hopfield network, which naturally incorporates symmetric token-/channel-mixing modules and layer normalization. Empirical studies reveal that symmetric interaction matrices in the model hinder performance in image recognition tasks. Introducing symmetry-breaking effects transitions the performance of the symmetric parallelized MLP-Mixer to that of the vanilla MLP-Mixer. This indicates that during standard training, weight matrices of the vanilla MLP-Mixer spontaneously acquire a symmetry-breaking configuration, enhancing their effectiveness. These findings offer insights into the intrinsic properties of Transformers and MLP-Mixers and their theoretical underpinnings, providing a robust framework for future model design and optimization.

cs.LG↗

arXiv2024

CAM Back Again: Large Kernel CNNs from a Weakly Supervised Object Localization Perspective

Recently, convolutional neural networks (CNNs) with large size kernels have attracted much attention in the computer vision field, following the success of the Vision Transformers. Large kernel CNNs have been reported to perform well in downstream vision tasks as well as in classification performance. The reason for the high-performance of large kernel CNNs in downstream tasks has been attributed to the large effective receptive field (ERF) produced by large size kernels, but this view has not been fully tested. We therefore revisit the performance of large kernel CNNs in downstream task, focusing on the weakly supervised object localization (WSOL) task. WSOL, a difficult downstream task that is not fully supervised, provides a new angle to explore the capabilities of the large kernel CNNs. Our study compares the modern large kernel CNNs ConvNeXt, RepLKNet, and SLaK to test the validity of the naive expectation that ERF size is important for improving downstream task performance. Our analysis of the factors contributing to high performance provides a different perspective, in which the main factor is feature map improvement. Furthermore, we find that modern CNNs are robust to the CAM problems of local regions of objects being activated, which has long been discussed in WSOL. CAM is the most classic WSOL method, but because of the above-mentioned problems, it is often used as a baseline method for comparison. However, experiments on the CUB-200-2011 dataset show that simply combining a large kernel CNN, CAM, and simple data augmentation methods can achieve performance (90.99% MaxBoxAcc) comparable to the latest WSOL method, which is CNN-based and requires special training or complex post-processing. The code is available at https://github.com/snskysk/CAM-Back-Again.

cs.CV↗

arXiv2023

iMixer: hierarchical Hopfield network implies an invertible, implicit and iterative MLP-Mixer

In the last few years, the success of Transformers in computer vision has stimulated the discovery of many alternative models that compete with Transformers, such as the MLP-Mixer. Despite their weak inductive bias, these models have achieved performance comparable to well-studied convolutional neural networks. Recent studies on modern Hopfield networks suggest the correspondence between certain energy-based associative memory models and Transformers or MLP-Mixer, and shed some light on the theoretical background of the Transformer-type architectures design. In this paper, we generalize the correspondence to the recently introduced hierarchical Hopfield network, and find iMixer, a novel generalization of MLP-Mixer model. Unlike ordinary feedforward neural networks, iMixer involves MLP layers that propagate forward from the output side to the input side. We characterize the module as an example of invertible, implicit, and iterative mixing module. We evaluate the model performance with various datasets on image classification tasks, and find that iMixer, despite its unique architecture, exhibits stable learning capabilities and achieves performance comparable to or better than the baseline vanilla MLP-Mixer. The results imply that the correspondence between the Hopfield networks and the Mixer models serves as a principle for understanding a broader class of Transformer-like architecture designs.

cs.LG↗

arXiv2023

FFT-based Dynamic Token Mixer for Vision

Multi-head-self-attention (MHSA)-equipped models have achieved notable performance in computer vision. Their computational complexity is proportional to quadratic numbers of pixels in input feature maps, resulting in slow processing, especially when dealing with high-resolution images. New types of token-mixer are proposed as an alternative to MHSA to circumvent this problem: an FFT-based token-mixer involves global operations similar to MHSA but with lower computational complexity. However, despite its attractive properties, the FFT-based token-mixer has not been carefully examined in terms of its compatibility with the rapidly evolving MetaFormer architecture. Here, we propose a novel token-mixer called Dynamic Filter and novel image recognition models, DFFormer and CDFFormer, to close the gaps above. The results of image classification and downstream tasks, analysis, and visualization show that our models are helpful. Notably, their throughput and memory efficiency when dealing with high-resolution image recognition is remarkable. Our results indicate that Dynamic Filter is one of the token-mixer options that should be seriously considered. The code is available at https://github.com/okojoalg/dfformer

cs.CV↗

arXiv2023

Example-Based Explainable AI and its Application for Remote Sensing Image Classification

We present a method of explainable artificial intelligence (XAI), "What I Know (WIK)", to provide additional information to verify the reliability of a deep learning model by showing an example of an instance in a training dataset that is similar to the input data to be inferred and demonstrate it in a remote sensing image classification task. One of the expected roles of XAI methods is verifying whether inferences of a trained machine learning model are valid for an application, and it is an important factor that what datasets are used for training the model as well as the model architecture. Our data-centric approach can help determine whether the training dataset is sufficient for each inference by checking the selected example data. If the selected example looks similar to the input data, we can confirm that the model was not trained on a dataset with a feature distribution far from the feature of the input data. With this method, the criteria for selecting an example are not merely data similarity with the input data but also data similarity in the context of the model task. Using a remote sensing image dataset from the Sentinel-2 satellite, the concept was successfully demonstrated with reasonably selected examples. This method can be applied to various machine-learning tasks, including classification and regression.

cs.AI↗

arXiv2022

Sequencer: Deep LSTM for Image Classification

In recent computer vision research, the advent of the Vision Transformer (ViT) has rapidly revolutionized various architectural design efforts: ViT achieved state-of-the-art image classification performance using self-attention found in natural language processing, and MLP-Mixer achieved competitive performance using simple multi-layer perceptrons. In contrast, several studies have also suggested that carefully redesigned convolutional neural networks (CNNs) can achieve advanced performance comparable to ViT without resorting to these new ideas. Against this background, there is growing interest in what inductive bias is suitable for computer vision. Here we propose Sequencer, a novel and competitive architecture alternative to ViT that provides a new perspective on these issues. Unlike ViTs, Sequencer models long-range dependencies using LSTMs rather than self-attention layers. We also propose a two-dimensional version of Sequencer module, where an LSTM is decomposed into vertical and horizontal LSTMs to enhance performance. Despite its simplicity, several experiments demonstrate that Sequencer performs impressively well: Sequencer2D-L, with 54M parameters, realizes 84.6% top-1 accuracy on only ImageNet-1K. Not only that, we show that it has good transferability and the robust resolution adaptability on double resolution-band.

cs.CV↗

arXiv2022

Application of QUBO solver using black-box optimization to structural design for resonance avoidance

Quadratic unconstrained binary optimization (QUBO) solvers can be applied to design an optimal structure to avoid resonance. QUBO algorithms that work on a classical or quantum device have succeeded in some industrial applications. However, their applications are still limited due to the difficulty of transforming from the original optimization problem to QUBO. Recently, black-box optimization (BBO) methods have been proposed to tackle this issue using a machine learning technique and a Bayesian treatment for combinatorial optimization. We employed the BBO methods to design a printed circuit board for resonance avoidance. This design problem is formulated to maximize natural frequency and simultaneously minimize the number of mounting points. The natural frequency, which is the bottleneck for the QUBO formulation, is approximated to a quadratic model in the BBO method. We demonstrated that BBO using a factorization machine shows good performance in both the calculation time and the success probability of finding the optimal solution. Our results can open up QUBO solvers' potential for other applications in structural designs.

cond-mat.mes-hall↗

arXiv2021

Counterfactual Explanation of Brain Activity Classifiers using Image-to-Image Transfer by Generative Adversarial Network

Deep neural networks (DNNs) can accurately decode task-related information from brain activations. However, because of the nonlinearity of the DNN, the decisions made by DNNs are hardly interpretable. One of the promising approaches for explaining such a black-box system is counterfactual explanation. In this framework, the behavior of a black-box system is explained by comparing real data and realistic synthetic data that are specifically generated such that the black-box system outputs an unreal outcome. Here we introduce a novel generative DNN (counterfactual activation generator, CAG) that can provide counterfactual explanations for DNN-based classifiers of brain activations. Importantly, CAG can simultaneously handle image transformation among multiple classes associated with different behavioral tasks. Using CAG, we demonstrated counterfactual explanation of DNN-based classifiers that learned to discriminate brain activations of seven behavioral tasks. Furthermore, by iterative applications of CAG, we were able to enhance and extract subtle spatial brain activity patterns that affected the classifier's decisions. Together, these results demonstrate that the counterfactual explanation based on image-to-image transformation would be a promising approach to understand and extend the current application of DNNs in fMRI analyses.

q-bio.NC↗

arXiv2021

RaftMLP: How Much Can Be Done Without Attention and with Less Spatial Locality?

For the past ten years, CNN has reigned supreme in the world of computer vision, but recently, Transformer has been on the rise. However, the quadratic computational cost of self-attention has become a serious problem in practice applications. There has been much research on architectures without CNN and self-attention in this context. In particular, MLP-Mixer is a simple architecture designed using MLPs and hit an accuracy comparable to the Vision Transformer. However, the only inductive bias in this architecture is the embedding of tokens. This leaves open the possibility of incorporating a non-convolutional (or non-local) inductive bias into the architecture, so we used two simple ideas to incorporate inductive bias into the MLP-Mixer while taking advantage of its ability to capture global correlations. A way is to divide the token-mixing block vertically and horizontally. Another way is to make spatial correlations denser among some channels of token-mixing. With this approach, we were able to improve the accuracy of the MLP-Mixer while reducing its parameters and computational complexity. The small model that is RaftMLP-S is comparable to the state-of-the-art global MLP-based model in terms of parameters and efficiency per calculation. In addition, we tackled the problem of fixed input image resolution for global MLP-based models by utilizing bicubic interpolation. We demonstrated that these models could be applied as the backbone of architectures for downstream tasks such as object detection. However, it did not have significant performance and mentioned the need for MLP-specific architectures for downstream tasks for global MLP-based models. The source code in PyTorch version is available at \url{https://github.com/okojoalg/raft-mlp}.

cs.CV↗

arXiv2021

Machine learning application to Fermi-LAT data: sharpening all-sky map and emphasizing variable sources

A novel application of machine-learning (ML) based image processing algorithms is proposed to analyze an all-sky map (ASM) obtained using the Fermi Gamma-ray Space Telescope. An attempt was made to simulate a one-year ASM from a short-exposure ASM generated from one-week observation by applying three ML based image processing algorithms: dictionary learning, U-net, and Noise2Noise. Although the inference based on ML is less clear compared to standard likelihood analysis, the quality of the ASM was generally improved. In particular, the complicated diffuse emission associated with the galactic plane was successfully reproduced only from one-week observation data to mimic a ground truth (GT) generated from a one-year observation. Such ML algorithms can be implemented relatively easily to provide sharper images without various assumptions of emission models. In contrast, large deviations between simulated ML maps and GT map were found, which are attributed to the significant temporal variability of blazar-type active galactic nuclei (AGNs) over a year. Thus, the proposed ML methods are viable not only to improve the image quality of an ASM, but also to detect variable sources, such as AGNs, algorithmically, i.e., without human bias. Moreover, we argue that this approach is widely applicable to ASMs obtained by various other missions; thus, it has the potential to examine giant structures and transient events, both of which are rarely found in pointing observations.

astro-ph.HE↗

arXiv2017

Deep Residual Networks and Weight Initialization

Residual Network (ResNet) is the state-of-the-art architecture that realizes successful training of really deep neural network. It is also known that good weight initialization of neural network avoids problem of vanishing/exploding gradients. In this paper, simplified models of ResNets are analyzed. We argue that goodness of ResNet is correlated with the fact that ResNets are relatively insensitive to choice of initial weights. We also demonstrate how batch normalization improves backpropagation of deep ResNets without tuning initial values of weights.

cs.LG↗

arXiv2015

More on 5d descriptions of 6d SCFTs

We propose new five-dimensional gauge theory descriptions of six-dimensional $\mathcal{N}=(1,0)$ superconformal field theories arising from type IIA brane configurations including an $ON^0$-plane. The new five-dimensional gauge theories may have $SO$, $Sp$, and $SU$ gauge groups and further broaden the landscape of ultraviolet complete five-dimensional $\mathcal{N}=1$ supersymmetric gauge theories. When we include an $O8^-$-plane in addition to an $ON^0$-plane, T-duality yields two $O7^-$-planes at the intersections of an $ON^0$-plane and two $O5^0$-planes. We propose a novel resolution of the $O7^-$-plane with four D7-branes in such a configuration, which enables us to obtain three different types of five-dimensional gauge theories, depending on whether we resolve either none or one or two $O7^-$-planes. Such different possibilities yield a new five-dimensional duality between a D-type $SU$ quiver and an $SO-Sp$ quiver theories. We also claim that a twisted circle compactification of a six-dimensional superconformal field theory may lead to a five-dimensional gauge theory different from those obtained by a simple circle compactification.

hep-th↗