×
前沿速递 | ncs 集萃：2026-10-01 期
朗读
总结
总结
字号
参考基因组和化石修订了蝙蝠家族系统发育和生物地理学
Reference genomes and fossils revise bat family phylogeny and biogeography
Bats are extraordinary among mammals, having uniquely evolved powered flight and laryngeal echolocation, along with disease resistance, extended healthspans and the ability to hibernate. However, the evolutionary history of bats and the understanding of these adaptations remain unresolved. We analysed chromosome-level, long-read genome assemblies from 103 bat species, including 42 new assemblies, representing all 21 bat families. This dataset, expanded in scope and assembly quality, yielded a ne
本研究通过分析103种蝙蝠（涵盖全部21科）的染色体水平长读长基因组组装数据，结合699个形态特征和44个前第四纪化石的整合分析，重构了蝙蝠系统发育树并揭示其适应性特征的演化历程：明确Myzopodidae为Vespertilionoidea最早分支、扬翅目内部关系及26条祖先染色体；通过化石化出生-死亡模型和扩散-灭绝分支分析表明蝙蝠及其动力飞行可能起源于晚古新世欧洲，且喉部回声定位早于冠群蝙蝠多样化；总证据定年法显著缩短了未被代表的基部分支长度，解决了长期争议的蝙蝠起源与适应性演化问题。
综合基因组与形态学数据证实蝙蝠动力飞行起源于晚古新世欧洲，喉部回声定位早于冠群分化，并重构了其染色体演化历史。
不完全信息博弈的可扩展决策
Scalable decision-making for games of imperfect information
Real-world decision-making generally involves hidden information, that is, information that is unknown to one agent but possessed by another. Unfortunately, the presence of large amounts of hidden information renders established reinforcement learning and search approaches ineffective. Even with multimillion-dollar industrial research efforts, top-human-level play at Stratego-a board wargame with hidden information on a massive scale-has remained beyond the reach of artificial intelligence (AI).
现实决策通常涉及隐藏信息，即一方未知而另一方掌握的信息；然而大量隐藏信息的存在使既有的强化学习和搜索方法失效，即便投入数百万美元的工业研究也未能让AI在具有海量隐藏信息的棋盘战争游戏Stratego中达到人类顶尖水平；本文提出基于自对弈强化学习与测试时搜索技术的通用型AI系统Ataraxos，其以显著优势击败史上最杰出的Stratego人类选手，实现该领域首个超人类结果，且计算资源与数据消耗较前序工作降低多个数量级；采用相同技术构建的Barrage Stratego、Hanabi和斗地主AI均展现超人类或领先性能，验证了该方法在对抗性、合作性与团队游戏中处理大规模隐藏信息的有效性，为战略决策领域提供了普适设计范式。
基于自对弈强化学习与测试时搜索的通用框架使AI在含海量隐藏信息的多类型游戏中实现超人类表现，突破传统方法局限并大幅降低计算成本。