General intelligence requires systems that acquire new skills efficiently and generalize beyond their training distributions. Although program synthesis approaches have strong generalization power, they face scaling issues due to the large combinatorial spaces that quickly render them impractical, requiring human-generated DSLs or pre-trained priors to narrow this search space. On the other hand, deep learning methods have had high successes, but they lack structured test-time adaptation and rely on heavy stochastic sampling or expensive gradient updates for fine-tuning. In this work, we propose the Latent Program Network (LPN), a novel architecture that builds in test-time search directly into neural models. LPN learns a latent space of implicit programs -- neurally mapping inputs to outputs -- through which it can search using gradients at test time. LPN combines the adaptability of symbolic approaches and the scalability of neural methods. It searches through a compact latent space at test time and bypasses the need for pre-defined domain-specific languages. On a range of programming-by-examples tasks, LPN either outperforms or matches performance compared to in-context learning and test-time training methods. Tested on the ARC-AGI benchmark, we demonstrate that LPN can both learn a compact program space and search through it at test time to adapt to novel tasks. LPN doubles its performance on out-of-distribution tasks when test-time search is switched on.
Searching for Quantum Effects in the Brain: A Bell-Type Test for Nonclassical Latent Representations in Autoencoders
School of Science, Zhejiang University of Science and Technology(浙江大学科学技术学院);Department of Physics and Institute of Theoretical and Computational Physics, University of Crete(希腊克里特大学物理系和理论与计算物理研究所);College of Life and Environmental Sciences, Hangzhou Normal University(杭州师范大学生命与环境科学学院);School of Automation and Electrical Engineering, Zhejiang University of Science and Technology(浙江大学科学技术学院自动化与电气工程学院);John A. Paulson School of Engineering and Applied Sciences, Harvard University(哈佛大学约翰·A·保罗森工程与应用科学学院)
Latent Chemical Space Searching for Plug-in Multi-objective Molecule Generation
State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University(北京大学药学院天然药物及仿生药物国家重点实验室);State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University(南京大学医药生物技术国家重点实验室)
Dublin Institute for Advanced Studies(都柏林高级研究院);California Institute of Technology(加州理工学院);Department of Physics, Northeastern University(东北大学物理系);Department of Mathematics, Stanford University(斯坦福大学数学系)
AI-driven Multimodal Representation Learning for Latent Mediation Structure Discovery of Socioeconomic Disadvantage, Psychosocial Factors, and Cardiometabolic Multimorbidity: Insights from the All of Us Research Program
Yale University(耶鲁大学)
本研究基于All of Us研究计划数据,开发AI驱动的多模态中介框架,通过变分自编码器推导潜在表征,发现心理社会脆弱性是连接社会经济劣势与心脏代谢共病的关键中介通路,为探究健康相关复杂关系提供了新方法。
Toward Compiler World Models: Learning Latent Dynamics for Efficient Tensor Program Search
Hangzhou Institute for Advanced Study, University of Chinese Academy of Sciences(中国科学院大学杭州高等研究院);Institute of Software, Chinese Academy of Sciences(中国科学院软件研究所);University of Chinese Academy of Sciences(中国科学院大学)
The weird and the wonderful in our Solar System: Searching for serendipity in the Legacy Survey of Space and Time
University of Oxford(牛津大学);Breakthrough Listen, University of California, Berkeley(加州大学伯克利分校突破聆听计划);SETI Institute(地外文明搜索研究所);University of Washington(华盛顿大学)