Pub Date : 2025-12-04DOI: 10.1038/s42256-025-01155-y
Binxu Wang, Carlos R. Ponce
Even before training, convolutional neural networks may reflect the brain’s visual processing principles. A study now shows how structure alone can help to explain the alignment between brains and models.
{"title":"Structure as an inductive bias for brain–model alignment","authors":"Binxu Wang, Carlos R. Ponce","doi":"10.1038/s42256-025-01155-y","DOIUrl":"10.1038/s42256-025-01155-y","url":null,"abstract":"Even before training, convolutional neural networks may reflect the brain’s visual processing principles. A study now shows how structure alone can help to explain the alignment between brains and models.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 12","pages":"1895-1896"},"PeriodicalIF":23.9,"publicationDate":"2025-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145680424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1038/s42256-025-01135-2
Chi-Hieu Nguyen, Thai Hoang Dinh, Diep N. Nguyen, Kristin Lauter, Miran Kim
Deep reinforcement learning (DRL) demonstrates significant potential in solving complex control and decision-making problems, but it may inadvertently expose sensitive, environment-specific information, raising privacy and security concerns for computer systems, humans and organizations. This work introduces a privacy-preserving framework using homomorphic encryption and advanced learning algorithms to secure DRL processes. Our framework enables the encryption of sensitive information, including states, actions and rewards, before sharing it with an untrusted processing platform. This encryption ensures data privacy, prevents unauthorized access and maintains compliance with data protection laws throughout the learning process. In addition, we develop innovative algorithms to efficiently handle a wide range of encrypted control tasks. Our core innovation is the homomorphic encryption-compatible Adam optimizer, which reparameterizes momentum values to bypass the need for high-degree polynomial approximations of inverse square roots on encrypted data. This adaptation, previously unexplored in homomorphic encryption-based ML research, enables stable and efficient training with adaptive learning rates in encrypted domains, addressing a critical bottleneck for privacy-preserving DRL with sparse rewards. Evaluations on standard DRL benchmarks demonstrate that our encrypted DRL performs comparably with its unencrypted counterpart (with a gap of less than 10%) and maintaining data confidentiality with homomorphic encryption. This work facilitates the integration of privacy-preserving DRL into real-world applications, addressing critical privacy concerns, and promoting the ethical advancement of artificial intelligence. A secure artificial intelligence framework is introduced that leverages homomorphic encryption to safeguard sensitive information in deep reinforcement learning, achieving accurate decision-making and ensuring data privacy and confidentiality.
{"title":"Empowering artificial intelligence with homomorphic encryption for secure deep reinforcement learning","authors":"Chi-Hieu Nguyen, Thai Hoang Dinh, Diep N. Nguyen, Kristin Lauter, Miran Kim","doi":"10.1038/s42256-025-01135-2","DOIUrl":"10.1038/s42256-025-01135-2","url":null,"abstract":"Deep reinforcement learning (DRL) demonstrates significant potential in solving complex control and decision-making problems, but it may inadvertently expose sensitive, environment-specific information, raising privacy and security concerns for computer systems, humans and organizations. This work introduces a privacy-preserving framework using homomorphic encryption and advanced learning algorithms to secure DRL processes. Our framework enables the encryption of sensitive information, including states, actions and rewards, before sharing it with an untrusted processing platform. This encryption ensures data privacy, prevents unauthorized access and maintains compliance with data protection laws throughout the learning process. In addition, we develop innovative algorithms to efficiently handle a wide range of encrypted control tasks. Our core innovation is the homomorphic encryption-compatible Adam optimizer, which reparameterizes momentum values to bypass the need for high-degree polynomial approximations of inverse square roots on encrypted data. This adaptation, previously unexplored in homomorphic encryption-based ML research, enables stable and efficient training with adaptive learning rates in encrypted domains, addressing a critical bottleneck for privacy-preserving DRL with sparse rewards. Evaluations on standard DRL benchmarks demonstrate that our encrypted DRL performs comparably with its unencrypted counterpart (with a gap of less than 10%) and maintaining data confidentiality with homomorphic encryption. This work facilitates the integration of privacy-preserving DRL into real-world applications, addressing critical privacy concerns, and promoting the ethical advancement of artificial intelligence. A secure artificial intelligence framework is introduced that leverages homomorphic encryption to safeguard sensitive information in deep reinforcement learning, achieving accurate decision-making and ensuring data privacy and confidentiality.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 12","pages":"1913-1926"},"PeriodicalIF":23.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145645246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-28DOI: 10.1038/s42256-025-01146-z
Daniel Durstewitz, Bruno Averbeck, Georgia Koppe
Modern artificial intelligence (AI) models, such as large language models, are usually trained once on a huge corpus of data, potentially fine-tuned for a specific task and then deployed with fixed parameters. Their training is costly, slow and gradual, requiring billions of repetitions. In stark contrast, animals continuously adapt to the ever-changing contingencies in their environments. This is particularly important for social species, where behavioural policies and reward outcomes may frequently change in interaction with peers. The underlying computational processes are often marked by rapid shifts in an animal’s behaviour and rather sudden transitions in neuronal population activity. Such computational capacities are of growing importance for AI systems operating in the real world, like those guiding robots or autonomous vehicles, or for agentic AI interacting with humans online. Can AI learn from neuroscience? This Perspective explores this question, integrating the literature on continual and in-context learning in AI with the neuroscience of learning on behavioural tasks with shifting rules, reward probabilities or outcomes. We outline an agenda for how the links between neuroscience and AI could be tightened, thus supporting the transfer of ideas and findings between both areas and contributing to the evolving field of NeuroAI. Durstewitz et al. explore what artificial intelligence can learn from the brain’s ability to adjust quickly to changing environments. By linking neuroscience studies of flexible behaviour with advances in continual and in-context learning, this Perspective outlines ways to strengthen the exchange of ideas between the two fields and advance NeuroAI.
{"title":"What neuroscience can tell AI about learning in continuously changing environments","authors":"Daniel Durstewitz, Bruno Averbeck, Georgia Koppe","doi":"10.1038/s42256-025-01146-z","DOIUrl":"10.1038/s42256-025-01146-z","url":null,"abstract":"Modern artificial intelligence (AI) models, such as large language models, are usually trained once on a huge corpus of data, potentially fine-tuned for a specific task and then deployed with fixed parameters. Their training is costly, slow and gradual, requiring billions of repetitions. In stark contrast, animals continuously adapt to the ever-changing contingencies in their environments. This is particularly important for social species, where behavioural policies and reward outcomes may frequently change in interaction with peers. The underlying computational processes are often marked by rapid shifts in an animal’s behaviour and rather sudden transitions in neuronal population activity. Such computational capacities are of growing importance for AI systems operating in the real world, like those guiding robots or autonomous vehicles, or for agentic AI interacting with humans online. Can AI learn from neuroscience? This Perspective explores this question, integrating the literature on continual and in-context learning in AI with the neuroscience of learning on behavioural tasks with shifting rules, reward probabilities or outcomes. We outline an agenda for how the links between neuroscience and AI could be tightened, thus supporting the transfer of ideas and findings between both areas and contributing to the evolving field of NeuroAI. Durstewitz et al. explore what artificial intelligence can learn from the brain’s ability to adjust quickly to changing environments. By linking neuroscience studies of flexible behaviour with advances in continual and in-context learning, this Perspective outlines ways to strengthen the exchange of ideas between the two fields and advance NeuroAI.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 12","pages":"1897-1912"},"PeriodicalIF":23.9,"publicationDate":"2025-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145611610","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-19DOI: 10.1038/s42256-025-01122-7
Simon Deltadahl, Julian Gilbey, Christine Van Laer, Nancy Boeckx, Mathie P. G. Leers, Tanya Freeman, Laura Aiken, Timothy Farren, Matthew Smith, Mohamad Zeina, Stephen MacDonald, Daniel Gleghorn, BloodCounts! consortium, James HF Rudd, Concetta Piazzese, Joseph Taylor, Nicholas Gleadall, Carola-Bibiane Schönlieb, Suthesh Sivapalaratnam, Michael Roberts, Parashkev Nachev
Blood cell morphology assessment via light microscopy constitutes a cornerstone of haematological diagnostics, providing crucial insights into diverse pathological conditions. This complex task demands expert interpretation owing to subtle morphological variations, biological heterogeneity and technical imaging factors that obstruct automated approaches. Conventional machine learning methods using discriminative models struggle with domain shifts, intraclass variability and rare morphological variants, constraining their clinical utility. We introduce CytoDiffusion, a diffusion-based generative classifier that faithfully models the distribution of blood cell morphology, combining accurate classification with robust anomaly detection, resistance to distributional shifts, interpretability, data efficiency and uncertainty quantification that surpasses clinical experts. Our approach outperforms state-of-the-art discriminative models in anomaly detection (area under the curve, 0.990 versus 0.916), resistance to domain shifts (0.854 versus 0.738 accuracy) and performance in low-data regimes (0.962 versus 0.924 balanced accuracy). In particular, CytoDiffusion generates synthetic blood cell images that expert haematologists cannot distinguish from real ones (accuracy, 0.523; 95% confidence interval: [0.505, 0.542]), demonstrating good command of the underlying distribution. Furthermore, we enhance model explainability through directly interpretable counterfactual heat maps. Our comprehensive evaluation framework establishes a multidimensional benchmark for medical image analysis in haematology, ultimately enabling improved diagnostic accuracy in clinical settings. Diffusion models are reframed by developing a generative blood cell classifier that performs reliably in low-data regimes, adapts to domain shifts, detects anomalies with robustness and provides uncertainty estimates that surpass clinical expert benchmarks.
通过光学显微镜进行血细胞形态评估是血液学诊断的基石，为不同的病理状况提供了重要的见解。由于细微的形态变化、生物异质性和技术成像因素阻碍了自动化方法，这项复杂的任务需要专家解释。使用判别模型的传统机器学习方法与领域转移、类内变异性和罕见的形态变异作斗争，限制了它们的临床应用。我们介绍了CytoDiffusion，这是一种基于扩散的生成分类器，它忠实地模拟了血细胞形态的分布，将准确的分类与鲁棒的异常检测、对分布变化的抵抗、可解释性、数据效率和不确定性量化相结合，超越了临床专家。我们的方法在异常检测（曲线下面积，0.990 vs 0.916）、抗域移（0.854 vs 0.738精度）和低数据状态下的性能（0.962 vs 0.924平衡精度）方面优于最先进的判别模型。特别是，CytoDiffusion生成的合成血细胞图像，血液专家无法将其与真实的血细胞图像区分开来（准确率为0.523；95%置信区间：[0.505,0.542]），显示出对潜在分布的良好掌握。此外，我们通过直接可解释的反事实热图增强了模型的可解释性。我们的综合评估框架建立了血液学医学图像分析的多维基准，最终提高了临床诊断的准确性。扩散模型通过开发再生血细胞分类器进行重构，该分类器在低数据状态下可靠地执行，适应域转移，鲁棒性检测异常，并提供超过临床专家基准的不确定性估计。
{"title":"Deep generative classification of blood cell morphology","authors":"Simon Deltadahl, Julian Gilbey, Christine Van Laer, Nancy Boeckx, Mathie P. G. Leers, Tanya Freeman, Laura Aiken, Timothy Farren, Matthew Smith, Mohamad Zeina, Stephen MacDonald, Daniel Gleghorn, BloodCounts! consortium, James HF Rudd, Concetta Piazzese, Joseph Taylor, Nicholas Gleadall, Carola-Bibiane Schönlieb, Suthesh Sivapalaratnam, Michael Roberts, Parashkev Nachev","doi":"10.1038/s42256-025-01122-7","DOIUrl":"10.1038/s42256-025-01122-7","url":null,"abstract":"Blood cell morphology assessment via light microscopy constitutes a cornerstone of haematological diagnostics, providing crucial insights into diverse pathological conditions. This complex task demands expert interpretation owing to subtle morphological variations, biological heterogeneity and technical imaging factors that obstruct automated approaches. Conventional machine learning methods using discriminative models struggle with domain shifts, intraclass variability and rare morphological variants, constraining their clinical utility. We introduce CytoDiffusion, a diffusion-based generative classifier that faithfully models the distribution of blood cell morphology, combining accurate classification with robust anomaly detection, resistance to distributional shifts, interpretability, data efficiency and uncertainty quantification that surpasses clinical experts. Our approach outperforms state-of-the-art discriminative models in anomaly detection (area under the curve, 0.990 versus 0.916), resistance to domain shifts (0.854 versus 0.738 accuracy) and performance in low-data regimes (0.962 versus 0.924 balanced accuracy). In particular, CytoDiffusion generates synthetic blood cell images that expert haematologists cannot distinguish from real ones (accuracy, 0.523; 95% confidence interval: [0.505, 0.542]), demonstrating good command of the underlying distribution. Furthermore, we enhance model explainability through directly interpretable counterfactual heat maps. Our comprehensive evaluation framework establishes a multidimensional benchmark for medical image analysis in haematology, ultimately enabling improved diagnostic accuracy in clinical settings. Diffusion models are reframed by developing a generative blood cell classifier that performs reliably in low-data regimes, adapts to domain shifts, detects anomalies with robustness and provides uncertainty estimates that surpass clinical expert benchmarks.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 11","pages":"1791-1803"},"PeriodicalIF":23.9,"publicationDate":"2025-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s42256-025-01122-7.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145547308","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The scalable solution to constrained combinatorial problems in high dimensions can address many challenges encountered in scientific and engineering disciplines. Inspired by the use of graph neural networks for quadratic-cost combinatorial optimization problems, Heydaribeni and colleagues proposed HypOp, which aims to efficiently solve general problems with higher-order constraints by leveraging hypergraph neural networks to extend previous algorithms to arbitrary cost functions. It incorporates a distributed training architecture to handle larger-scale tasks efficiently. Here we reproduce the primary experiments of HypOp and examine its robustness with respect to the number of graphics processing units, distributed partitioning strategies and fine-tuning methods. We also assess its transferability by applying it to the maximum clique problem and the quadratic assignment problem. The results validate the reusability of HypOp across diverse application scenarios. Furthermore, we provide guidelines offering practical insights for effectively applying it to multiple combinatorial optimization problems. HypOp is a scalable method for solving complex combinatorial problems. This study reproduces its results, tests its robustness, extends it to new tasks and provides practical guidelines for broader scientific applications.
{"title":"Reusability report: A distributed strategy for solving combinatorial optimization problems with hypergraph neural networks","authors":"Xiaodi Li, Jianfeng Gui, Wei Xue, Baochuan Wang, Kai Chen, Pijing Wei, Junfeng Xia, Zhenyu Yue","doi":"10.1038/s42256-025-01141-4","DOIUrl":"10.1038/s42256-025-01141-4","url":null,"abstract":"The scalable solution to constrained combinatorial problems in high dimensions can address many challenges encountered in scientific and engineering disciplines. Inspired by the use of graph neural networks for quadratic-cost combinatorial optimization problems, Heydaribeni and colleagues proposed HypOp, which aims to efficiently solve general problems with higher-order constraints by leveraging hypergraph neural networks to extend previous algorithms to arbitrary cost functions. It incorporates a distributed training architecture to handle larger-scale tasks efficiently. Here we reproduce the primary experiments of HypOp and examine its robustness with respect to the number of graphics processing units, distributed partitioning strategies and fine-tuning methods. We also assess its transferability by applying it to the maximum clique problem and the quadratic assignment problem. The results validate the reusability of HypOp across diverse application scenarios. Furthermore, we provide guidelines offering practical insights for effectively applying it to multiple combinatorial optimization problems. HypOp is a scalable method for solving complex combinatorial problems. This study reproduces its results, tests its robustness, extends it to new tasks and provides practical guidelines for broader scientific applications.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 11","pages":"1870-1878"},"PeriodicalIF":23.9,"publicationDate":"2025-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145547309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Translating mass spectra into chemical structures is a central challenge in exposomics, making it difficult to quickly track the millions of chemicals found in humans and the environment. Unlike metabolomics, key problems in developing models for chemicals with a larger molecular space include data scarcity, model complexity and proper query strategy. Here we present a molecular structure generator (MSGo) that can generate structures directly from mass spectra and discover unknown polyfluorinated chemicals in the exposome. Trained with only virtual spectra using a transformer neural network, MSGo correctly identified 48% of structures in a validation set and was better at discovering new polyfluorinated chemicals in wastewater samples reported in the literature than experts. Applying probability-oriented masking to the virtual spectra is key to MSGo’s performance. Rapid discovery of chemicals with limited experimental mass spectral data using automated tools such as MSGo is key to tackling the current unknown polyfluorinated chemical crisis. Yu and colleagues present MSGo, an artificial intelligence exposomics tool trained on virtual mass spectra with masking that identifies pollutants by generating chemical structures that match measured spectral data.
{"title":"Pseudodata-based molecular structure generator to reveal unknown chemicals","authors":"Nanyang Yu, Zheng Ma, Qi Shao, Laihui Li, Xuebing Wang, Bingcai Pan, Hongxia Yu, Si Wei","doi":"10.1038/s42256-025-01140-5","DOIUrl":"10.1038/s42256-025-01140-5","url":null,"abstract":"Translating mass spectra into chemical structures is a central challenge in exposomics, making it difficult to quickly track the millions of chemicals found in humans and the environment. Unlike metabolomics, key problems in developing models for chemicals with a larger molecular space include data scarcity, model complexity and proper query strategy. Here we present a molecular structure generator (MSGo) that can generate structures directly from mass spectra and discover unknown polyfluorinated chemicals in the exposome. Trained with only virtual spectra using a transformer neural network, MSGo correctly identified 48% of structures in a validation set and was better at discovering new polyfluorinated chemicals in wastewater samples reported in the literature than experts. Applying probability-oriented masking to the virtual spectra is key to MSGo’s performance. Rapid discovery of chemicals with limited experimental mass spectral data using automated tools such as MSGo is key to tackling the current unknown polyfluorinated chemical crisis. Yu and colleagues present MSGo, an artificial intelligence exposomics tool trained on virtual mass spectra with masking that identifies pollutants by generating chemical structures that match measured spectral data.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 11","pages":"1879-1887"},"PeriodicalIF":23.9,"publicationDate":"2025-11-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145509633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-13DOI: 10.1038/s42256-025-01145-0
Kristian Kersting
A new benchmark, KaBLE (knowledge and belief language evaluation), indicates that some large language models are unable to accurately distinguish belief from knowledge and fact, calling into question their use in real-word applications such as medicine and law.
{"title":"Large language models still struggle with false beliefs","authors":"Kristian Kersting","doi":"10.1038/s42256-025-01145-0","DOIUrl":"10.1038/s42256-025-01145-0","url":null,"abstract":"A new benchmark, KaBLE (knowledge and belief language evaluation), indicates that some large language models are unable to accurately distinguish belief from knowledge and fact, calling into question their use in real-word applications such as medicine and law.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 11","pages":"1778-1779"},"PeriodicalIF":23.9,"publicationDate":"2025-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145498438","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-13DOI: 10.1038/s42256-025-01143-2
Bradley H. Theilman, James B. Aimone
The finite element method (FEM) is one of the most important and ubiquitous numerical methods for solving partial differential equations (PDEs) on computers for scientific and engineering discovery. Applying the FEM to larger and more detailed scientific models has driven advances in high-performance computing for decades. Here we demonstrate that scalable spiking neuromorphic hardware can directly implement the FEM by constructing a spiking neural network that solves the large, sparse, linear systems of equations at the core of the FEM. We show that for the Poisson equation, a fundamental PDE in science and engineering, our neural circuit achieves meaningful levels of numerical accuracy and close to ideal scaling on modern, inherently parallel and energy-efficient neuromorphic hardware, specifically Intel’s Loihi 2 neuromorphic platform. We illustrate extensions to irregular mesh geometries in both two and three dimensions as well as other PDEs such as linear elasticity. Our spiking neural network is constructed from a recurrent network model of the brain’s motor cortex and, in contrast to black-box deep artificial neural network-based methods for PDEs, directly translates the well-understood and trusted mathematics of the FEM to a natively spiking neuromorphic algorithm. Theilman and Aimone introduce a natively spiking algorithm for solving partial differential equations on large-scale neuromorphic computers and demonstrate the algorithm on Intel’s Loihi 2 neuromorphic research chip.
{"title":"Solving sparse finite element problems on neuromorphic hardware","authors":"Bradley H. Theilman, James B. Aimone","doi":"10.1038/s42256-025-01143-2","DOIUrl":"10.1038/s42256-025-01143-2","url":null,"abstract":"The finite element method (FEM) is one of the most important and ubiquitous numerical methods for solving partial differential equations (PDEs) on computers for scientific and engineering discovery. Applying the FEM to larger and more detailed scientific models has driven advances in high-performance computing for decades. Here we demonstrate that scalable spiking neuromorphic hardware can directly implement the FEM by constructing a spiking neural network that solves the large, sparse, linear systems of equations at the core of the FEM. We show that for the Poisson equation, a fundamental PDE in science and engineering, our neural circuit achieves meaningful levels of numerical accuracy and close to ideal scaling on modern, inherently parallel and energy-efficient neuromorphic hardware, specifically Intel’s Loihi 2 neuromorphic platform. We illustrate extensions to irregular mesh geometries in both two and three dimensions as well as other PDEs such as linear elasticity. Our spiking neural network is constructed from a recurrent network model of the brain’s motor cortex and, in contrast to black-box deep artificial neural network-based methods for PDEs, directly translates the well-understood and trusted mathematics of the FEM to a natively spiking neuromorphic algorithm. Theilman and Aimone introduce a natively spiking algorithm for solving partial differential equations on large-scale neuromorphic computers and demonstrate the algorithm on Intel’s Loihi 2 neuromorphic research chip.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 11","pages":"1845-1857"},"PeriodicalIF":23.9,"publicationDate":"2025-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s42256-025-01143-2.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145498437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-13DOI: 10.1038/s42256-025-01144-1
Mohammad Nadeem, Shahab Saquib Sohail, Erik Cambria, Shagufta Afreen
Biases in artificial intelligence models have been studied predominantly through Western lenses, overlooking South Asia’s unique contexts of caste, religion, colourism and representation. This Comment highlights region-specific biases in language and vision models and calls for fairness frameworks grounded in South Asian realities.
{"title":"South Asian biases in language and vision models","authors":"Mohammad Nadeem, Shahab Saquib Sohail, Erik Cambria, Shagufta Afreen","doi":"10.1038/s42256-025-01144-1","DOIUrl":"10.1038/s42256-025-01144-1","url":null,"abstract":"Biases in artificial intelligence models have been studied predominantly through Western lenses, overlooking South Asia’s unique contexts of caste, religion, colourism and representation. This Comment highlights region-specific biases in language and vision models and calls for fairness frameworks grounded in South Asian realities.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 11","pages":"1775-1777"},"PeriodicalIF":23.9,"publicationDate":"2025-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145498436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-13DOI: 10.1038/s42256-025-01142-3
Atlas Kazemian, Eric Elmoznino, Michael F. Bonner
What underlies the emergence of cortex-aligned representations in deep neural network models of vision? Earlier work suggested that shared architectural constraints were a major factor, but the success of widely varied architectures after pretraining raises critical questions about the importance of architectural constraints. Here we show that in wide networks with minimal training, architectural inductive biases have a prominent role. We examined networks with varied architectures but no pretraining and quantified their ability to predict image representations in the visual cortices of monkeys and humans. We found that cortex-aligned representations emerge in convolutional architectures that combine two key manipulations of dimensionality: compression in the spatial domain, through pooling, and expansion in the feature domain by increasing the number of channels. We further show that the inductive biases of convolutional architectures are critical for obtaining performance gains from feature expansion—dimensionality manipulations were relatively ineffective in other architectures and in convolutional models with targeted lesions. Our findings suggest that the architectural constraints of convolutional networks are sufficiently close to the constraints of biological vision to allow many aspects of cortical visual representation to emerge even before synaptic connections have been tuned through experience. Kazemian et al. report that untrained convolutional networks with wide layers predict primate visual cortex responses nearly as well as task-optimized networks, revealing how architectural constraints shape brain-like representations in deep networks.
{"title":"Convolutional architectures are cortex-aligned de novo","authors":"Atlas Kazemian, Eric Elmoznino, Michael F. Bonner","doi":"10.1038/s42256-025-01142-3","DOIUrl":"10.1038/s42256-025-01142-3","url":null,"abstract":"What underlies the emergence of cortex-aligned representations in deep neural network models of vision? Earlier work suggested that shared architectural constraints were a major factor, but the success of widely varied architectures after pretraining raises critical questions about the importance of architectural constraints. Here we show that in wide networks with minimal training, architectural inductive biases have a prominent role. We examined networks with varied architectures but no pretraining and quantified their ability to predict image representations in the visual cortices of monkeys and humans. We found that cortex-aligned representations emerge in convolutional architectures that combine two key manipulations of dimensionality: compression in the spatial domain, through pooling, and expansion in the feature domain by increasing the number of channels. We further show that the inductive biases of convolutional architectures are critical for obtaining performance gains from feature expansion—dimensionality manipulations were relatively ineffective in other architectures and in convolutional models with targeted lesions. Our findings suggest that the architectural constraints of convolutional networks are sufficiently close to the constraints of biological vision to allow many aspects of cortical visual representation to emerge even before synaptic connections have been tuned through experience. Kazemian et al. report that untrained convolutional networks with wide layers predict primate visual cortex responses nearly as well as task-optimized networks, revealing how architectural constraints shape brain-like representations in deep networks.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"7 11","pages":"1834-1844"},"PeriodicalIF":23.9,"publicationDate":"2025-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145498435","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：