Nature Machine Intelligence最新文献_第3页

Recognizing reproducibility and reusability in times of fast science 认识到在快速科学时代的再现性和可重用性

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-25 DOI: 10.1038/s42256-026-01219-7

引用次数: 0

Computational framework to predict and shape human–machine interactions in closed-loop, co-adaptive neural interfaces 在闭环、共适应神经接口中预测和塑造人机交互的计算框架

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-23 DOI: 10.1038/s42256-026-01194-z

Maneeshika M. Madduri, Momona Yamagami, Si Jia Li, Sasha Burckhardt, Samuel A. Burden, Amy L. Orsborn

{"title":"Computational framework to predict and shape human–machine interactions in closed-loop, co-adaptive neural interfaces","authors":"Maneeshika M. Madduri, Momona Yamagami, Si Jia Li, Sasha Burckhardt, Samuel A. Burden, Amy L. Orsborn","doi":"10.1038/s42256-026-01194-z","DOIUrl":"10.1038/s42256-026-01194-z","url":null,"abstract":"Neural interfaces can restore or augment human sensorimotor capabilities by converting high-bandwidth biological signals into control signals for an external device via a decoder algorithm. Leveraging user and decoder adaptation to create co-adaptive interfaces presents opportunities to improve usability and personalize devices. However, we lack principled methods to model and optimize the complex two-learner dynamics that arise in co-adaptive interfaces. Here we present computational methods based on control theory and game theory to analyse and generate predictions for user–decoder co-adaptive outcomes in continuous interactions. We tested these computational methods using an experimental platform in which human participants (N = 14) learn to control a cursor using an adaptive myoelectric interface to track a target on a computer display. Our framework allowed us to characterize user and decoder changes within co-adaptive myoelectric interfaces. Our framework further allowed us to predict how changes in the decoder algorithm impacted co-adaptive interface performance and revealed how interface properties can shape user behaviour. Our findings demonstrate an experimentally validated computational framework that can be used to design user–decoder interactions in closed-loop, co-adaptive neural interfaces. This framework opens future opportunities to optimize co-adaptive neural interfaces to expand the performance and application domains for neural interfaces. Madduri et al. introduce a computational framework grounded in control and game theory to model co-adaptation between users and decoders in neural interfaces. This framework enables a principled design of closed-loop systems that improve usability and personalization.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"8 3","pages":"372-387"},"PeriodicalIF":23.9,"publicationDate":"2026-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s42256-026-01194-z.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147496843","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}

引用次数: 0

Interpretability and implicit model semantics in biomedicine and deep learning 生物医学和深度学习中的可解释性和隐式模型语义

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-23 DOI: 10.1038/s42256-026-01177-0

Jonathan Warrell, Michael Gancz, Hussein Mohsen, Prashant Emani, Mark Gerstein

引用次数: 0

LLMs displaying less cognitive bias are not necessarily better decision makers 表现出较少认知偏见的法学硕士不一定是更好的决策者

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-17 DOI: 10.1038/s42256-026-01208-w

Vittoria Dentella, Marco Marelli, Luca Rinaldi

引用次数: 0

Sample-efficient generative molecular design using memory manipulation 使用记忆操作的样本高效生成分子设计

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-17 DOI: 10.1038/s42256-026-01200-4

Jeff Guo, Junwu Chen, Anthony GX-Chen, Philippe Schwaller

{"title":"Sample-efficient generative molecular design using memory manipulation","authors":"Jeff Guo, Junwu Chen, Anthony GX-Chen, Philippe Schwaller","doi":"10.1038/s42256-026-01200-4","DOIUrl":"10.1038/s42256-026-01200-4","url":null,"abstract":"Generative molecular design for drug discovery has recently achieved a wave of experimental validation. Language models operating on string-based representations of molecules are amongst the most successful architectures. The most important factor for downstream success is whether an in silico oracle (computational predictor of a molecule property) is well correlated with the desired end point (such as binding affinity). To this end, current methods use cheaper proxy oracles with a higher throughput before evaluating the most promising subset with high-fidelity oracles. The ability to directly generate molecules with optimal properties as predicted by high-fidelity oracles (computationally expensive simulations with greater predictive accuracy) could greatly enhance generative design and improve hit rates. However, current models are not efficient enough to consider such a prospect, exemplifying the sample efficiency problem. Recently, the Mamba architecture has been proposed as an alternative to transformers, which are widely used in large language models. Existing works have validated Mamba’s performance on tasks spanning natural language completion to biology foundation models. In this work, we introduce a framework called Saturn, which demonstrates the application of the Mamba architecture for generative molecular design. Here we elucidate how experience replay with data augmentation improves the sample efficiency and how Mamba intensifies the effect of this mechanism. Next, we show that Mamba with experience replay outperforms 16 models on multiparameter optimization tasks relevant to drug discovery and possesses sufficient sample efficiency to directly optimize density functional theory simulations as a high-fidelity oracle. Guo et al. train a Mamba-based language model for molecule generation and find that data augmentation and experience replay can enable the efficient generation of property-optimized small molecules.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"8 3","pages":"449-460"},"PeriodicalIF":23.9,"publicationDate":"2026-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147465106","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}

引用次数: 0

A robot operating system framework for using large language models in embodied AI 在嵌入式人工智能中使用大型语言模型的机器人操作系统框架

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-16 DOI: 10.1038/s42256-026-01186-z

Christopher E. Mower, Yuhui Wan, Hongzhan Yu, Antoine Grosnit, Jonas Gonzalez-Billandon, Matthieu Zimmer, Puze Liu, Daniel Palenicek, Davide Tateo, Jan Peters, Kaixian Qu, Mike Zhang, Guowei Lan, Andrei Cramariuc, Cesar Cadena, Marco Hutter, Guangjian Tian, Yuzhen Zhuang, Kun Shao, Xingyue Quan, Jianye Hao, Jun Wang, Haitham Bou-Ammar

{"title":"A robot operating system framework for using large language models in embodied AI","authors":"Christopher E. Mower, Yuhui Wan, Hongzhan Yu, Antoine Grosnit, Jonas Gonzalez-Billandon, Matthieu Zimmer, Puze Liu, Daniel Palenicek, Davide Tateo, Jan Peters, Kaixian Qu, Mike Zhang, Guowei Lan, Andrei Cramariuc, Cesar Cadena, Marco Hutter, Guangjian Tian, Yuzhen Zhuang, Kun Shao, Xingyue Quan, Jianye Hao, Jun Wang, Haitham Bou-Ammar","doi":"10.1038/s42256-026-01186-z","DOIUrl":"10.1038/s42256-026-01186-z","url":null,"abstract":"Autonomous robots capable of turning natural-language instructions into reliable physical actions remain a central challenge in artificial intelligence. Here we show that connecting a large language model agent to the robot operating system enables a versatile framework for embodied intelligence, and we release the complete implementation as freely available open-source code. The agent automatically translates large language model outputs into robot actions, supports interchangeable execution modes (inline code or behaviour trees), learns new atomic skills via imitation, and continually refines them through automated optimization and reflection from human or environmental feedback. Extensive experiments validate the framework, showcasing robustness, scalability and versatility in diverse scenarios and embodiments, including long-horizon tasks, tabletop rearrangements, dynamic task optimization and remote supervisory control. Moreover, all the results presented in this work were achieved by utilizing open-source pretrained large language models. Mower, Wan et al. introduce ROS-LLM, an open-source system that lets non-experts control robots with natural language, learn new skills from demonstrations and feedback, and automatically tune actions for reliable performance in real-world tasks.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"8 3","pages":"313-325"},"PeriodicalIF":23.9,"publicationDate":"2026-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147465107","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}

引用次数: 0

The case for stakeholder-driven AI auditing in automatic speech recognition 自动语音识别中利益相关者驱动的人工智能审计案例

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-16 DOI: 10.1038/s42256-026-01207-x

Mona Sloane, Hilke Schellmann, Katelyn Xiaoying Mei, Anna Seo Gyeong Choi, Allison Koenecke

引用次数: 0

Predicting and interpreting cell-type-specific drug responses in the small-data regime using inductive priors 利用诱导先验预测和解释小数据下细胞类型特异性药物反应

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-16 DOI: 10.1038/s42256-026-01202-2

Reem Alsulami, Robert Lehmann, Sumeer A. Khan, Vincenzo Lagani, Alberto Maillo, David Gomez-Cabrero, Narsis A. Kiani, Jesper Tegner

{"title":"Predicting and interpreting cell-type-specific drug responses in the small-data regime using inductive priors","authors":"Reem Alsulami, Robert Lehmann, Sumeer A. Khan, Vincenzo Lagani, Alberto Maillo, David Gomez-Cabrero, Narsis A. Kiani, Jesper Tegner","doi":"10.1038/s42256-026-01202-2","DOIUrl":"10.1038/s42256-026-01202-2","url":null,"abstract":"Predicting how small molecules affect diverse cell types phenotypically is central to drug discovery, yet it remains a challenging task. Modelling cell-type-specific transcriptional responses provides a scalable alternative for early candidate identification, enabling broader exploration and lower costs than exhaustive experimental exploration of the chemical space. Here we present PrePR-CT, a graph-based deep learning approach that utilizes cell-type-specific co-expression networks as an inductive bias to predict transcriptional responses to chemical perturbations. Graph attention networks learn biologically meaningful representations that capture cell-type-specific gene interactions, enabling gene-level attributions. Across five single-cell RNA sequencing datasets, including human blood and multiple cancer lines, one bulk transcriptomics dataset and a large-scale small-molecule screen, the method generalizes to unseen perturbations and previously unseen cell types under data-limited settings, achieving higher accuracy for expression variability compared to generative baselines. Attribution analyses identify high-attention genes that complement traditional differential expression analyses, highlighting pathway-specific mechanisms of small-molecule response. By combining scalability, robustness to distribution shifts and interpretability, PrePR-CT enables cell-type-resolved prediction of drug responses, providing a foundation for more precise modelling of cellular perturbations in early drug discovery. Alsulami et al. present PrePR-CT, a computational approach that predicts how different cell types respond to drug-like compounds using limited data. By integrating biological networks with machine learning, it improves accuracy, interpretability and efficiency in early drug discovery.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"8 3","pages":"461-473"},"PeriodicalIF":23.9,"publicationDate":"2026-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s42256-026-01202-2.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147465192","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}

引用次数: 0

A benchmarking framework for embodied neuromorphic agents 具身神经形态因子的基准框架

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-11 DOI: 10.1038/s42256-026-01197-w

Giulia D’Angelo, Jens E. Pedersen, Taimoor Hassan, Matteo Cianchetti, Josh Bongard, Fumiya Iida, Giacomo Indiveri, Matej Hoffmann, Cecilia Laschi, Chiara De Luca, Chiara Bartolozzi, Elisa Donati

{"title":"A benchmarking framework for embodied neuromorphic agents","authors":"Giulia D’Angelo, Jens E. Pedersen, Taimoor Hassan, Matteo Cianchetti, Josh Bongard, Fumiya Iida, Giacomo Indiveri, Matej Hoffmann, Cecilia Laschi, Chiara De Luca, Chiara Bartolozzi, Elisa Donati","doi":"10.1038/s42256-026-01197-w","DOIUrl":"10.1038/s42256-026-01197-w","url":null,"abstract":"Enabling robots to swiftly, robustly and efficiently interact with a dynamic environment remains a key challenge. The robotic community can draw inspiration from the co-adaptation and synergistic interplay between animals’ brains and bodies, which underpins embodied intelligence. Soft robots and neuromorphic technology offer a natural solution for such a challenge, enabling low-power, material-based and event-driven sensorimotor processing and control that seamlessly handles the continuous dynamic demands of embodied agents. In this Perspective, we propose a comprehensive framework for benchmarking neuromorphic computing (brain) that control soft robots (body), based on a suite of tasks, essential metrics and a reproducible robotic platform. The goal is to allow researchers to evaluate their embodied neuromorphic system with a physical robot, in real-world scenarios. The robotic platform is accessible, open-source, modular and scalable, so task complexity can be gradually increased, fostering a standardized approach. By coupling metrics with physical implementations, this framework will drive progress in soft robotics, neuromorphic computing and embodied intelligence. Combining soft robotics with neuromorphic engineering is a promising approach in embodied intelligence. Giulia d’Angelo et al. contribute to progress in this field by developing a framework for benchmarking neuromorphic controllers on soft robotic platforms.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"8 3","pages":"300-312"},"PeriodicalIF":23.9,"publicationDate":"2026-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147570394","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}

引用次数: 0

Learning collision risk proactively from naturalistic driving data at scale 从大规模的自然驾驶数据中主动学习碰撞风险

IF 23.9 1区计算机科学

Nature Machine Intelligence Pub Date : 2026-03-09 DOI: 10.1038/s42256-026-01189-w

Yiru Jiao (, ), Simeon C. Calvert, Sander van Cranenburgh, Hans van Lint

{"title":"Learning collision risk proactively from naturalistic driving data at scale","authors":"Yiru Jiao \u0000 (, ), Simeon C. Calvert, Sander van Cranenburgh, Hans van Lint","doi":"10.1038/s42256-026-01189-w","DOIUrl":"10.1038/s42256-026-01189-w","url":null,"abstract":"Accurately and proactively alerting drivers or automated systems to emerging collisions is crucial for road safety, particularly in highly interactive and complex urban environments. Existing methods require labour-intensive annotation of sparse risk, struggle to consider varying contextual factors or are tailored to limited scenarios. Here we present the generalized surrogate safety measure (GSSM), a data-driven approach that learns collision risk from naturalistic driving without the need for crash or risk labels. Trained on diverse datasets and evaluated on 2,591 real-world crashes and near-crashes, a basic GSSM using only instantaneous motion kinematics achieves an area under the precision–recall curve of 0.9 and secures a median time advance of 2.6 s to prevent potential collisions. Incorporating more interaction patterns and contextual factors provides further performance gains. Across interaction scenarios, such as rear end, merging and turning, GSSM consistently outperforms existing baselines in terms of accuracy and timeliness. These results establish GSSM as a scalable, context-aware and generalizable foundation for identifying risky interactions before they become unavoidable and support proactive safety in autonomous driving systems and traffic incident management. Jiao et al. introduce a generalized safety measure for autonomous driving systems that learns collision risk from everyday driving without labels. It accurately warns in real time of crashes and near-crashes and secures time for an early reaction.","PeriodicalId":48533,"journal":{"name":"Nature Machine Intelligence","volume":"8 3","pages":"337-350"},"PeriodicalIF":23.9,"publicationDate":"2026-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147381754","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}

引用次数: 0