Non-invasive brain stimulation techniques(NIBS),including repetitive transcranial magnetic stimulation(rTMS) and transcranial electric stim ulation(tES),are increasingly being adopted clinically for treatment of neuro...Non-invasive brain stimulation techniques(NIBS),including repetitive transcranial magnetic stimulation(rTMS) and transcranial electric stim ulation(tES),are increasingly being adopted clinically for treatment of neuropsychiatric and neurological disorders,albeit with varying success.The rationale behind the use of NIBS has historically been that stim ulation techniques modulate neuronal activity in the targeted region and consequently induce plasticity which can lead to therapeutic outcomes.展开更多
Brain damage sustained from repeated blows in boxing, wrestling, and other combat sports has serious physical and mental health consequences. The degenerative brain disease, chronic traumatic encephalopathy (CTE), pre...Brain damage sustained from repeated blows in boxing, wrestling, and other combat sports has serious physical and mental health consequences. The degenerative brain disease, chronic traumatic encephalopathy (CTE), presents clinically with memory loss, aggression, difficulty in rational thinking and other cognitive problems. This spectrum, which mimics Alzheimer’s disease, is diagnosed post-mortem through a brain biopsy in many professional athletes. However, little is known about the process of development and how to identify vulnerable individuals who may be on course for developing CTE. Boxing is a sport that has a severe toll on athletes’ health, primarily on their brain health and function. This review addresses the concerns of brain injury, describes the pathologies that manifest in multiple scales, e.g., molecular and cognitive, and also proposes possible diagnostic and prognostic markers to characterize the early onset of CTE along with the aim to identify a starting point for future precautions and interventions.展开更多
Few multi-agent reinforcement learning (MARL) researches on Google research football (GRF) focus on the 11-vs-11 multi-agent full-game scenario and to the best of our knowledge, no open benchmark on this scenario has ...Few multi-agent reinforcement learning (MARL) researches on Google research football (GRF) focus on the 11-vs-11 multi-agent full-game scenario and to the best of our knowledge, no open benchmark on this scenario has been released to the public. In this work, we fill the gap by providing a population-based MARL training pipeline and hyperparameter settings on multi-agent football scenario that outperforms the bot with difficulty 1.0 from scratch within 2 million steps. Our experiments serve as a reference for the expected performance of independent proximal policy optimization (IPPO), a state-of-the-art multi-agent reinforcement learning algorithm where each agent tries to maximize its own policy independently across various training configurations. Meanwhile, we release our training framework Light-MALib which extends the MALib codebase by distributed and asynchronous implementation with additional analytical tools for football games. Finally, we provide guidance for building strong football AI with population-based training and release diverse pretrained policies for benchmarking. The goal is to provide the community with a head start for whoever experiment their works on GRF and a simple-to-use population-based training framework for further improving their agents through self-play. The implementation is available at https://github.com/Shanghai-Digital-Brain-Laboratory/DB-Football.展开更多
Transformer architectures have facilitated the development of large-scale and general-purpose sequence models for prediction tasks in natural language processing and computer vision,e.g.,GPT-3 and Swin Transformer.Alt...Transformer architectures have facilitated the development of large-scale and general-purpose sequence models for prediction tasks in natural language processing and computer vision,e.g.,GPT-3 and Swin Transformer.Although originally designed for prediction problems,it is natural to inquire about their suitability for sequential decision-making and reinforcement learning problems,which are typically beset by long-standing issues involving sample efficiency,credit assignment,and partial observability.In recent years,sequence models,especially the Transformer,have attracted increasing interest in the RL communities,spawning numerous approaches with notable effectiveness and generalizability.This survey presents a comprehensive overview of recent works aimed at solving sequential decision-making tasks with sequence models such as the Transformer,by discussing the connection between sequential decision-making and sequence modeling,and categorizing them based on the way they utilize the Transformer.Moreover,this paper puts forth various potential avenues for future research intending to improve the effectiveness of large sequence models for sequential decision-making,encompassing theoretical foundations,network architectures,algorithms,and efficient training systems.展开更多
Main text The disruption of protein folding homeostasis in motoneurons(MNs)and the accumulation of protein aggregates are some of the main molecular hallmarks of amyotrophic lateral sclerosis(ALS).Evidence from sporad...Main text The disruption of protein folding homeostasis in motoneurons(MNs)and the accumulation of protein aggregates are some of the main molecular hallmarks of amyotrophic lateral sclerosis(ALS).Evidence from sporadic and familial ALS(fALS)patients and from ALS models suggests that protein aggregation directly participates in neurodegeneration.In turn,the loss of MN homeostasis triggers a coping mechanism,the integrated stress response(ISR)[1].The ISR is initiated by four independent stress-sensing kinases,each of them activated by distinct stresses:protein kinase R(PKR)by double-strand RNA,protein kinase RNA-like endoplasmic reticulum kinase(PERK)by protein misfolding at the endoplasmic reticulum(ER),general control nonderepressible 2(GCN2)by nutrient starvation,and heme-regulated inhibitor(HRI)by heme deprivation.展开更多
基金supported by the Bryant Stokes Neurological Research Fund (to JM)a fellowship from Multiple Sclerosis Western Australia (MSWA)+1 种基金the Perron Institute for Neurological and Translational Sciencethe Bryant Stokes Neurological Research Fund (to JR)。
文摘Non-invasive brain stimulation techniques(NIBS),including repetitive transcranial magnetic stimulation(rTMS) and transcranial electric stim ulation(tES),are increasingly being adopted clinically for treatment of neuropsychiatric and neurological disorders,albeit with varying success.The rationale behind the use of NIBS has historically been that stim ulation techniques modulate neuronal activity in the targeted region and consequently induce plasticity which can lead to therapeutic outcomes.
文摘Brain damage sustained from repeated blows in boxing, wrestling, and other combat sports has serious physical and mental health consequences. The degenerative brain disease, chronic traumatic encephalopathy (CTE), presents clinically with memory loss, aggression, difficulty in rational thinking and other cognitive problems. This spectrum, which mimics Alzheimer’s disease, is diagnosed post-mortem through a brain biopsy in many professional athletes. However, little is known about the process of development and how to identify vulnerable individuals who may be on course for developing CTE. Boxing is a sport that has a severe toll on athletes’ health, primarily on their brain health and function. This review addresses the concerns of brain injury, describes the pathologies that manifest in multiple scales, e.g., molecular and cognitive, and also proposes possible diagnostic and prognostic markers to characterize the early onset of CTE along with the aim to identify a starting point for future precautions and interventions.
基金supported by the National Natural Science Foundation of China(No.62206289).
文摘Few multi-agent reinforcement learning (MARL) researches on Google research football (GRF) focus on the 11-vs-11 multi-agent full-game scenario and to the best of our knowledge, no open benchmark on this scenario has been released to the public. In this work, we fill the gap by providing a population-based MARL training pipeline and hyperparameter settings on multi-agent football scenario that outperforms the bot with difficulty 1.0 from scratch within 2 million steps. Our experiments serve as a reference for the expected performance of independent proximal policy optimization (IPPO), a state-of-the-art multi-agent reinforcement learning algorithm where each agent tries to maximize its own policy independently across various training configurations. Meanwhile, we release our training framework Light-MALib which extends the MALib codebase by distributed and asynchronous implementation with additional analytical tools for football games. Finally, we provide guidance for building strong football AI with population-based training and release diverse pretrained policies for benchmarking. The goal is to provide the community with a head start for whoever experiment their works on GRF and a simple-to-use population-based training framework for further improving their agents through self-play. The implementation is available at https://github.com/Shanghai-Digital-Brain-Laboratory/DB-Football.
基金The SJTU team was partially supported by“New Generation of AI 2030”Major Project(2018AAA0100900)Shanghai Municipal Science and Technology Major Project(2021SHZDZX0102)+1 种基金the National Natural Science Foundation of China(Grant No.62076161)Muning Wen is supported by Wu Wen Jun Honorary Scholarship,AI Institute,Shanghai Jiao Tong University.
文摘Transformer architectures have facilitated the development of large-scale and general-purpose sequence models for prediction tasks in natural language processing and computer vision,e.g.,GPT-3 and Swin Transformer.Although originally designed for prediction problems,it is natural to inquire about their suitability for sequential decision-making and reinforcement learning problems,which are typically beset by long-standing issues involving sample efficiency,credit assignment,and partial observability.In recent years,sequence models,especially the Transformer,have attracted increasing interest in the RL communities,spawning numerous approaches with notable effectiveness and generalizability.This survey presents a comprehensive overview of recent works aimed at solving sequential decision-making tasks with sequence models such as the Transformer,by discussing the connection between sequential decision-making and sequence modeling,and categorizing them based on the way they utilize the Transformer.Moreover,this paper puts forth various potential avenues for future research intending to improve the effectiveness of large sequence models for sequential decision-making,encompassing theoretical foundations,network architectures,algorithms,and efficient training systems.
基金supported by PID2020-120497RB-I00 MCIU/AEI/https://doi.org/10.13039/501100011033,BFU2017-90043-P MCINN/AEI/https://doi.org/10.13039/501100011033/by FEDER“Una manera de hacer Europa”(MA and TA),Proyecto Intramural IdisNa 2022(MA),Fundación para la Investigación Médica Aplicada(FIMA)Proyectos I+D,2017(TA)and Fundación Occident and DalecandELA Association(MA)supported by República de Panamá,Programa de Becas IFARHU-SENACYT(reference number 270-2018-922),NP by AC FIMA pre-doctoral fellowship.
文摘Main text The disruption of protein folding homeostasis in motoneurons(MNs)and the accumulation of protein aggregates are some of the main molecular hallmarks of amyotrophic lateral sclerosis(ALS).Evidence from sporadic and familial ALS(fALS)patients and from ALS models suggests that protein aggregation directly participates in neurodegeneration.In turn,the loss of MN homeostasis triggers a coping mechanism,the integrated stress response(ISR)[1].The ISR is initiated by four independent stress-sensing kinases,each of them activated by distinct stresses:protein kinase R(PKR)by double-strand RNA,protein kinase RNA-like endoplasmic reticulum kinase(PERK)by protein misfolding at the endoplasmic reticulum(ER),general control nonderepressible 2(GCN2)by nutrient starvation,and heme-regulated inhibitor(HRI)by heme deprivation.
