Yongping Zhang, Shuting Jiang, Fuhong He, Yuanyuan Tian, Haiyang Hu, Li Gao, Lin Zhang, Aili Chen, Yixin Hu, Liyan Fan, Chun Yang, Bi Zhou, Dan Liu, Zihan Zhou, Yanxun Su, Lei Qin, Yi Wang, Hailong He, Jun Lu, Peifang Xiao, Shaoyan Hu, Qian-Fei Wang <<<

Abstract Background Cancer patients can achieve dramatic responses to chemotherapy yet retain resistant tumor cells, which ultimately results in relapse. Although xenograft model studies have identified several cellular and molecular features that are associated with chemoresistance in acute myeloid leukemia (AML), to what extent AML patients exhibit these properties remains largely unknown. Results We apply single-cell RNA sequencing to paired pre- and post-chemotherapy whole bone marrow samples obtained from 13 pediatric AML patients who had achieved disease remission, and distinguish AML clusters from normal cells based on their unique transcriptomic profiles. Approximately 50% of leukemic stem and progenitor populations actively express leukemia stem cell (LSC) and oxidative phosphorylation (OXPHOS) signatures, respectively. These clusters have a higher chance of tolerating therapy and exhibit an enhanced metabolic program in response to treatment. Interestingly, the transmembrane receptor CD69 is highly expressed in chemoresistant hematopoietic stem cell (HSC)-like populations (named the CD69+ HSC-like subpopulation). Furthermore, overexpression of CD69 results in suppression of the mTOR signaling pathway and promotion of cell quiescence and adhesion in vitro. Finally, the presence of CD69+ HSC-like cells is associated with unfavorable genetic mutations, the persistence of residual tumor cells in chemotherapy, and poor outcomes in independent pediatric and adult public AML cohorts. Conclusions Our analysis reveals leukemia stem cell and OXPHOS as two major chemoresistant features in human AML patients. CD69 may serve as a potential biomarker in defining a subpopulation of chemoresistant leukemia stem cells. These findings have important implications for targeting residual chemo-surviving AML cells. <<<

Alexander Haglund, Verena Zuber, Maya Abouzeid, Yifei Yang, Jeong Hun Ko, Liv Wiemann, Maria Otero-Jimenez, Louwai Muhammed, Rahel Feleke, Alexi Nott, James D. Mills, Liisi Laaniste, Djordje O. Gveric, Daniel Clode, Ann C. Babtie, Susanna Pagni, Ravishankara Bellampalli, Alyma Somani, Karina McDade, Jasper J. Anink, Lucia Mesarosova, Nurun Fancy, Nanet Willumsen, Amy Smith, Johanna Jackson, Javier Alegre-Abarrategui, Eleonora Aronica, Paul M. Matthews, Maria Thom, Sanjay M. Sisodiya, Prashant K. Srivastava, Dheeraj Malhotra, Julien Bryois, Leonardo Bottolo, Michael R. Johnson <<<

Abstract Gene expression quantitative trait loci are widely used to infer relationships between genes and central nervous system (CNS) phenotypes; however, the effect of brain disease on these inferences is unclear. Using 2,348,438 single-nuclei profiles from 391 disease-case and control brains, we report 13,939 genes whose expression correlated with genetic variation, of which 16.7–40.8% (depending on cell type) showed disease-dependent allelic effects. Across 501 colocalizations for 30 CNS traits, 23.6% had a disease dependency, even after adjusting for disease status. To estimate the unconfounded effect of genes on outcomes, we repeated the analysis using nondiseased brains (n = 183) and reported an additional 91 colocalizations not present in the larger mixed disease and control dataset, demonstrating enhanced interpretation of disease-associated variants. Principled implementation of single-cell Mendelian randomization in control-only brains identified 140 putatively causal gene–trait associations, of which 11 were replicated in the UK Biobank, prioritizing candidate peripheral biomarkers predictive of CNS outcomes. <<<

DeepSeek-AI, Aixin Liu, Bei Feng, Bin Wang, Bingxuan Wang, Bo Liu, Chenggang Zhao, Chengqi Dengr, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fuli Luo, Guangbo Hao, Guanting Chen, Guowei Li, H. Zhang, Hanwei Xu, Hao Yang, Haowei Zhang, Honghui Ding, Huajian Xin, Huazuo Gao, Hui Li, Hui Qu, J. L. Cai, Jian Liang, Jianzhong Guo, Jiaqi Ni, Jiashi Li, Jin Chen, Jingyang Yuan, Junjie Qiu, Junxiao Song, Kai Dong, Kaige Gao, Kang Guan, Lean Wang, Lecong Zhang, Lei Xu, Leyi Xia, Liang Zhao, Liyue Zhang, Meng Li, Miaojun Wang, Mingchuan Zhang, Minghua Zhang, Minghui Tang, Mingming Li, Ning Tian, Panpan Huang, Peiyi Wang, Peng Zhang, Qihao Zhu, Qinyu Chen, Qiushi Du, R. J. Chen, R. L. Jin, Ruiqi Ge, Ruizhe Pan, Runxin Xu, Ruyi Chen, S. S. Li, Shanghao Lu, Shangyan Zhou, Shanhuang Chen, Shaoqing Wu, Shengfeng Ye, Shirong Ma, Shiyu Wang, Shuang Zhou, Shuiping Yu, Shunfeng Zhou, Size Zheng, T. Wang, Tian Pei, Tian Yuan, Tianyu Sun, W. L. Xiao, Wangding Zeng, Wei An, Wen Liu, Wenfeng Liang, Wenjun Gao, Wentao Zhang, X. Q. Li, Xiangyue Jin, Xianzu Wang, Xiao Bi, Xiaodong Liu, Xiaohan Wang, Xiaojin Shen, Xiaokang Chen, Xiaosha Chen, Xiaotao Nie, Xiaowen Sun, Xiaoxiang Wang, Xin Liu, Xin Xie, Xingkai Yu, Xinnan Song, Xinyi Zhou, Xinyu Yang, Xuan Lu, Xuecheng Su, Y. Wu, Y. K. Li, Y. X. Wei, Y. X. Zhu, Yanhong Xu, Yanping Huang, Yao Li, Yao Zhao, Yaofeng Sun, Yaohui Li, Yaohui Wang, Yi Zheng, Yichao Zhang, Yiliang Xiong, Yilong Zhao, Ying He, Ying Tang, Yishi Piao, Yixin Dong, Yixuan Tan, Yiyuan Liu, Yongji Wang, Yongqiang Guo, Yuchen Zhu, Yuduan Wang, Yuheng Zou, Yukun Zha, Yunxian Ma, Yuting Yan, Yuxiang You, Yuxuan Liu, Z. Z. Ren, Zehui Ren, Zhangli Sha, Zhe Fu, Zhen Huang, Zhen Zhang, Zhenda Xie, Zhewen Hao, Zhihong Shao, Zhiniu Wen, Zhipeng Xu, Zhongyu Zhang, Zhuoshu Li, Zihan Wang, Zihui Gu, Zilin Li, Ziwei Xie <<<

We present DeepSeek-V2, a strong Mixture-of-Experts (MoE) language modelcharacterized by economical training and efficient inference. It comprises 236Btotal parameters, of which 21B are activated for each token, and supports acontext length of 128K tokens. DeepSeek-V2 adopts innovative architecturesincluding Multi-head Latent Attention (MLA) and DeepSeekMoE. MLA guaranteesefficient inference through significantly compressing the Key-Value (KV) cacheinto a latent vector, while DeepSeekMoE enables training strong models at aneconomical cost through sparse computation. Compared with DeepSeek 67B,DeepSeek-V2 achieves significantly stronger performance, and meanwhile saves42.5% of training costs, reduces the KV cache by 93.3%, and boosts the maximumgeneration throughput to 5.76 times. We pretrain DeepSeek-V2 on a high-qualityand multi-source corpus consisting of 8.1T tokens, and further performSupervised Fine-Tuning (SFT) and Reinforcement Learning (RL) to fully unlockits potential. Evaluation results show that, even with only 21B activatedparameters, DeepSeek-V2 and its chat versions still achieve top-tierperformance among open-source models. <<<

Caihua Sang, Qiang Liu, Yiwei Lai, Shijun Xia, Ruhong Jiang, Songnan Li, Qi Guo, Qifan Li, Mingyang Gao, Xueyuan Guo, Lihong Huang, Nian Liu, Chenxi Jiang, Song Zuo, Xiaoxia Liu, Mengmeng Li, Weili Ge, Shangming Song, Lianghua Chen, Shuanglun Xie, Jiangang Zou, Ke Chen, Xiangfei Liu, Hesheng Hu, Xinhua Wang, Jinlin Zhang, Zhaojun Wang, Chi Wang, Liu He, Chao Jiang, Ribo Tang, Ning Zhou, Yunlong Wang, Deyong Long, Xin Du, Chenyang Jiang, Laurent Macle, Jianzeng Dong, Changsheng Ma, , Wei Wang, Xin Zhao, Changyi Li, Zhuo Liang, Xu Li, Xiangyi Kong, Wenli Dai, Yufeng Wang, Lu Yu, Xueyan Ding, Hui Cheng, Jianwei Lin, Pei Zhang, Yaxun Sun, Xiaofeng Hou, Yao Wang, Yumei Xue, Rong Bai, Jing Du <<<

ImportanceSuccess rates of pulmonary vein isolation (PVI) are modest for persistent atrial fibrillation (AF). Additional linear ablation beyond PVI has not been proved superior to PVI alone in randomized trials. Ethanol infusion of the vein of Marshall (EIVOM) facilitates ablation at the mitral isthmus and may lead to improved effectiveness of a linear ablation strategy.ObjectiveTo determine whether linear ablation with radiofrequency energy combined with EIVOM added to PVI improves sinus rhythm maintenance compared with PVI alone in patients with persistent AF.Design, Setting, and ParticipantsThe PROMPT-AF trial is an investigator-initiated, multicenter, open-label, randomized trial involving 12 tertiary hospitals in China. A total of 498 patients aged 18 to 80 years, with AF persisting for more than 3 months, undergoing first-time AF ablation, were enrolled and randomized from August 27, 2021, to July 16, 2023.InterventionsPatients were randomized to undergo PVI alone or PVI plus EIVOM and linear ablation (intervention). The latter group first underwent EIVOM, followed by PVI and linear ablation of the left atrial roof, mitral isthmus, and cavotricuspid isthmus.Main Outcomes and MeasuresThe primary end point was freedom from any documented atrial arrhythmias lasting more than 30 seconds, without the use of antiarrhythmic drugs within 12 months. Secondary outcomes included freedom from atrial arrhythmia recurrence, AF, atrial arrhythmia recurrence after multiple procedures, and documented atrial tachycardia or atrial flutter with or without antiarrhythmic drugs; AF burden; and improvement in quality of life. Patients were monitored with wearable single-lead electrocardiographic (ECG) patches, worn for 24 hours a week, supplemented by symptom-triggered ECGs and Holter monitoring.ResultsAmong 498 randomized patients, 495 (99.4%) were included in the primary analysis (mean age, 61.1 years [SD, 9.7] years, 361 male [72.9%]). After 12 months, 174 of 246 patients (70.7%) assigned to undergo PVI plus EIVOM and linear ablation and 153 of 249 patients (61.5%) assigned to undergo PVI alone remained free from atrial arrhythmias without taking antiarrhythmic drugs (hazard ratio, 0.73; 95% CI, 0.54-0.99, P = .045). The intervention effect was consistent across all prespecified subgroups. The comparison of secondary outcomes did not demonstrate significant results.ConclusionAmong patients with persistent AF, linear ablation combined with EIVOM in addition to PVI significantly improved freedom from atrial arrhythmias within 12 months compared with PVI alone.Trial RegistrationClinicalTrials.gov Identifier: NCT04497376 <<<

NP-hard problems regularly come up in video games, with interestingconnections to real-world problems. In the game Minecraft, players placetorches on the ground to light up dark areas. Placing them in a way thatminimizes the total number of torches to save resources is far from trivial. Inthis paper, we use Quantum Computing to approach this problem. To this end, wederive a QUBO formulation of the torch placement problem, which we uncover tobe very similar to another NP-hard problem. We employ a solution strategy thatinvolves learning Lagrangian weights in an iterative process, adding to theever growing toolbox of QUBO formulations. Finally, we perform experiments onreal quantum hardware using real game data to demonstrate that our approachyields good torch placements. <<<

Mohammed Toufiq, Darawan Rinchai, Eleonore Bettacchioli, Basirudeen Syed Ahamed Kabeer, Taushif Khan, Bishesh Subba, Olivia White, Marina Yurieva, Joshy George, Noemie Jourde-Chiche, Laurent Chiche, Karolina Palucka, Damien Chaussabel <<<

AbstractBackgroundFeature selection is a critical step for translating advances afforded by systems-scale molecular profiling into actionable clinical insights. While data-driven methods are commonly utilized for selecting candidate genes, knowledge-driven methods must contend with the challenge of efficiently sifting through extensive volumes of biomedical information. This work aimed to assess the utility of large language models (LLMs) for knowledge-driven gene prioritization and selection.MethodsIn this proof of concept, we focused on 11 blood transcriptional modules associated with an Erythroid cells signature. We evaluated four leading LLMs across multiple tasks. Next, we established a workflow leveraging LLMs. The steps consisted of: (1) Selecting one of the 11 modules; (2) Identifying functional convergences among constituent genes using the LLMs; (3) Scoring candidate genes across six criteria capturing the gene’s biological and clinical relevance; (4) Prioritizing candidate genes and summarizing justifications; (5) Fact-checking justifications and identifying supporting references; (6) Selecting a top candidate gene based on validated scoring justifications; and (7) Factoring in transcriptome profiling data to finalize the selection of the top candidate gene.ResultsOf the four LLMs evaluated, OpenAI's GPT-4 and Anthropic's Claude demonstrated the best performance and were chosen for the implementation of the candidate gene prioritization and selection workflow. This workflow was run in parallel for each of the 11 erythroid cell modules by participants in a data mining workshop. Module M9.2 served as an illustrative use case. The 30 candidate genes forming this module were assessed, and the top five scoring genes were identified as BCL2L1, ALAS2, SLC4A1, CA1, and FECH. Researchers carefully fact-checked the summarized scoring justifications, after which the LLMs were prompted to select a top candidate based on this information. GPT-4 initially chose BCL2L1, while Claude selected ALAS2. When transcriptional profiling data from three reference datasets were provided for additional context, GPT-4 revised its initial choice to ALAS2, whereas Claude reaffirmed its original selection for this module.ConclusionsTaken together, our findings highlight the ability of LLMs to prioritize candidate genes with minimal human intervention. This suggests the potential of this technology to boost productivity, especially for tasks that require leveraging extensive biomedical knowledge. <<<

AbstractCancer diagnosis and therapeutics (theranostics) based on the tumor microenvironment (TME) and biomarkers has been an emerging approach for precision medicine. DNA nanotechnology dynamically controls the self‐assembly of DNA molecules at the nanometer scale to construct intelligent DNA chemical reaction systems. The DNA logic circuit is a particularly emerging approach for computing within the DNA chemical systems. DNA logic circuits can sensitively respond to tumor‐specific markers and the TME through logic operations and signal amplification, to generate detectable signals or to release anti‐cancer agents. In this review, the fundamental concepts of DNA logic circuits are clarified, the basic modules in the circuit are summarized, and how this advanced nano‐assembly circuit responds to tumor‐related molecules, how to perform logic operations, to realize signal amplification, and selectively release drugs through discussing over 30 application examples, are demonstrated. This review shows that DNA logic circuits have powerful logic judgment and signal amplification functions in improving the specificity and sensitivity of cancer diagnosis and making cancer treatment controllable. In the future, researchers are expected to overcome the existing shortcomings of DNA logic circuits and design smarter DNA devices with better biocompatibility and stability, which will further promote the development of cancer theranostics. <<<

Abstract The societal shift toward greater gender equality has led to increased variability in people’s gender role attitudes, or the belief that men and women should occupy distinct family roles (i.e. men as breadwinners and women as homemakers). Existing evidence on the association between gender role attitudes and relationship well-being remains inconclusive with mixed findings, likely because past research has not adequately considered the direction and degree of (in)congruencies between partners within the relationship. Using longitudinal samples of 1,327 couples from the United States and 5,856 couples from Germany tracked over 2 and 13 years, respectively, we employed dyadic response surface analysis to examine how different patterns of partner (in)congruencies in gender role attitudes predict relationship well-being in mixed-gender relationships. The results showed that, for US men and German men and women, the direction of incongruence between partners’ gender role attitudes mattered: relationship satisfaction was higher when men adopted more egalitarian attitudes than women (or conversely, when women adopted more traditional attitudes than men) compared with the reverse. Relationship satisfaction was also higher when both partners showed congruence in extreme gender role attitudes (either strongly traditional or egalitarian) than when either partner endorsed more neutral attitudes. US women reported higher relationship satisfaction only when either partner endorsed more egalitarian attitudes. Although past research emphasizes the benefits of partner similarity for relationship well-being, our findings highlight the importance of both similarity and complementarity in gender role attitudes, potentially subject to cultural and contextual factors. <<<

This article is about the neural conundrum behind the slowness of humanbehavior. The information throughput of a human being is about 10 bits/s. Incomparison, our sensory systems gather data at ~10^9 bits/s. The stark contrastbetween these numbers remains unexplained and touches on fundamental aspects ofbrain function: What neural substrate sets this speed limit on the pace of ourexistence? Why does the brain need billions of neurons to process 10 bits/s?Why can we only think about one thing at a time? The brain seems to operate intwo distinct modes: the "outer" brain handles fast high-dimensional sensory andmotor signals, whereas the "inner" brain processes the reduced few bits neededto control behavior. Plausible explanations exist for the large neuron numbersin the outer brain, but not for the inner brain, and we propose new researchdirections to remedy this. <<<

Yangyang Shao, Ning Lu, Zhenfang Wu, Chen Cai, Shanshan Wang, Ling-Li Zhang, Fan Zhou, Shijun Xiao, Lin Liu, Xiaofei Zeng, Huajun Zheng, Chen Yang, Zhihu Zhao, Guoping Zhao, Jin-Qiu Zhou, Xiaoli Xue, Zhongjun Qin <<<

Xin Wei, Mengjiao Chen, Qi Zhang, Junyi Gong, Jie Liu, Kaicheng Yong, Qin Wang, Jiongjiong Fan, Suhui Chen, Hua Hua, Zhaowei Luo, Xiaoyan Zhao, Xuan Wang, Wei Li, Jia Cong, Xiting Yu, Zhihan Wang, Ruipeng Huang, Jiaxin Chen, Xiaoyi Zhou, Jie Qiu, Ping Xu, Jeremy Murray, Hai Wang, Yang Xu, Chenwu Xu, Gen Xu, Jinliang Yang, Bin Han, Xuehui Huang <<<

Understanding how numerous quantitative trait loci (QTL) shape phenotypic variation is an important question in genetics. To address this, we established a permanent population of 18,421 (18K) rice lines with reduced population structure. We generated reference-level genome assemblies of the founders and genotyped all 18K-rice lines through whole-genome sequencing. Through high-resolution mapping, 96 high-quality candidate genes contributing to variation in 16 traits were identified, including OsMADS22 and OsFTL1 verified as causal genes for panicle number and heading date, respectively. We identified epistatic QTL pairs and constructed a genetic interaction network with 19 genes serving as hubs. Overall, 170 masking epistasis pairs were characterized, serving as an important factor contributing to genetic background effects across diverse varieties. The work provides a basis to guide grain yield and quality improvements in rice. <<<

AbstractRest and sleep not only strengthen existing memories but also reorganise memories to generate new knowledge that extends beyond direct experience. However, it remains unclear bothhowmemories are reorganised and the effect of this reorganisation on behaviour. Here, we designed a novel protocol to casually manipulate memory consolidation during rest using awake, contextual targeted memory reactivation (TMR). We found that promoting memory consolidation during rest qualitatively reorganises memories byforming ‘shortcuts’ between memorieswhich have not been experienced together. These shortcuts in memory extend beyond direct experience to facilitate our ability to make novel inferences. A series of control tests indicate that inference performance cannot be explained by quantitative strengthening of the experienced component links but are rather explained by qualitative changes in the cognitive map which involve formation of new shortcuts. Interestingly, we show that representing a shortcut may come with limitations, as shortcuts cannot be readily updated in response to rapid changes in the environment. Together, these findings reveal how memories are reorganised during awake rest to construct a cognitive map of our environment, while highlighting the constraints set by a trade-off between efficient and flexible behaviour. <<<

Foivos Georgiadis, Sara Larivière, David Glahn, L. Elliot Hong, Peter Kochunov, Bryan Mowry, Carmel Loughland, Christos Pantelis, Frans A. Henskens, Melissa J. Green, Murray J. Cairns, Patricia T. Michie, Paul E. Rasser, Stanley Catts, Paul Tooney, Rodney J. Scott, Ulrich Schall, Vaughan Carr, Yann Quidé, Axel Krug, Frederike Stein, Igor Nenadić, Katharina Brosch, Tilo Kircher, Raquel Gur, Ruben Gur, Theodore D. Satterthwaite, Andriana Karuk, Edith Pomarol- Clotet, Joaquim Radua, Paola Fuentes-Claramonte, Raymond Salvador, Gianfranco Spalletta, Aristotle Voineskos, Kang Sim, Benedicto Crespo-Facorro, Diana Tordesillas Gutiérrez, Stefan Ehrlich, Nicolas Crossley, Dominik Grotegerd, Jonathan Repple, Rebekka Lencer, Udo Dannlowski, Vince Calhoun, Kelly Rootes-Murdy, Caroline Demro, Ian S. Ramsay, Scott R. Sponheim, Andre Schmidt, Stefan Borgwardt, Alexander Tomyshev, Irina Lebedeva, Cyril Höschl, Filip Spaniel, Adrian Preda, Dana Nguyen, Anne Uhlmann, Dan J. Stein, Fleur Howells, Henk S. Temmingh, Ana M. Diaz Zuluaga, Carlos López Jaramillo, Felice Iasevoli, Ellen Ji, Stephanie Homan, Wolfgang Omlor, Philipp Homan, Stefan Kaiser, Erich Seifritz, Bratislav Misic, Sofie L. Valk, Paul Thompson, Theo G. M. van Erp, Jessica A. Turner, , Boris Bernhardt, Matthias Kirschner <<<

AbstractSchizophrenia is a prototypical network disorder with widespread brain-morphological alterations, yet it remains unclear whether these distributed alterations robustly reflect the underlying network layout. We tested whether large-scale structural alterations in schizophrenia relate to normative structural and functional connectome architecture, and systematically evaluated robustness and generalizability of these network-level alterations. Leveraging anatomical MRI scans from 2439 adults with schizophrenia and 2867 healthy controls from 26 ENIGMA sites and normative data from the Human Connectome Project (n = 207), we evaluated structural alterations of schizophrenia against two network susceptibility models: (i) hub vulnerability, which examines associations between regional network centrality and magnitude of disease-related alterations; (ii) epicenter mapping, which identifies regions whose typical connectivity profile most closely resembles the disease-related morphological alterations. To assess generalizability and specificity, we contextualized the influence of site, disease stages, and individual clinical factors and compared network associations of schizophrenia with that found in affective disorders. Our findings show schizophrenia-related cortical thinning is spatially associated with functional and structural hubs, suggesting that highly interconnected regions are more vulnerable to morphological alterations. Predominantly temporo-paralimbic and frontal regions emerged as epicenters with connectivity profiles linked to schizophrenia’s alteration patterns. Findings were robust across sites, disease stages, and related to individual symptoms. Moreover, transdiagnostic comparisons revealed overlapping epicenters in schizophrenia and bipolar, but not major depressive disorder, suggestive of a pathophysiological continuity within the schizophrenia-bipolar-spectrum. In sum, cortical alterations over the course of schizophrenia robustly follow brain network architecture, emphasizing marked hub susceptibility and temporo-frontal epicenters at both the level of the group and the individual. Subtle variations of epicenters across disease stages suggest interacting pathological processes, while associations with patient-specific symptoms support additional inter-individual variability of hub vulnerability and epicenters in schizophrenia. Our work outlines potential pathways to better understand macroscale structural alterations, and inter- individual variability in schizophrenia. <<<

Weixin Liang, Zachary Izzo, Yaohui Zhang, Haley Lepp, Hancheng Cao, Xuandong Zhao, Lingjiao Chen, Haotian Ye, Sheng Liu, Zhi Huang, Daniel A. McFarland, James Y. Zou <<<

We present an approach for estimating the fraction of text in a large corpuswhich is likely to be substantially modified or produced by a large languagemodel (LLM). Our maximum likelihood model leverages expert-written andAI-generated reference texts to accurately and efficiently examine real-worldLLM-use at the corpus level. We apply this approach to a case study ofscientific peer review in AI conferences that took place after the release ofChatGPT: ICLR 2024, NeurIPS 2023, CoRL 2023 and EMNLP 2023. Our results suggestthat between 6.5% and 16.9% of text submitted as peer reviews to theseconferences could have been substantially modified by LLMs, i.e. beyondspell-checking or minor writing updates. The circumstances in which generatedtext occurs offer insight into user behavior: the estimated fraction ofLLM-generated text is higher in reviews which report lower confidence, weresubmitted close to the deadline, and from reviewers who are less likely torespond to author rebuttals. We also observe corpus-level trends in generatedtext which may be too subtle to detect at the individual level, and discuss theimplications of such trends on peer review. We call for futureinterdisciplinary work to examine how LLM use is changing our information andknowledge practices. <<<

This paper shows that masked autoencoders (MAE) are scalable self-supervisedlearners for computer vision. Our MAE approach is simple: we mask randompatches of the input image and reconstruct the missing pixels. It is based ontwo core designs. First, we develop an asymmetric encoder-decoder architecture,with an encoder that operates only on the visible subset of patches (withoutmask tokens), along with a lightweight decoder that reconstructs the originalimage from the latent representation and mask tokens. Second, we find thatmasking a high proportion of the input image, e.g., 75%, yields a nontrivialand meaningful self-supervisory task. Coupling these two designs enables us totrain large models efficiently and effectively: we accelerate training (by 3xor more) and improve accuracy. Our scalable approach allows for learninghigh-capacity models that generalize well: e.g., a vanilla ViT-Huge modelachieves the best accuracy (87.8%) among methods that use only ImageNet-1Kdata. Transfer performance in downstream tasks outperforms supervisedpre-training and shows promising scaling behavior. <<<

Arun C. Habermann, Austin J. Gutierrez, Linh T. Bui, Stephanie L. Yahn, Nichelle I. Winters, Carla L. Calvi, Lance Peter, Mei-I Chung, Chase J. Taylor, Christopher Jetter, Latha Raju, Jamie Roberson, Guixiao Ding, Lori Wood, Jennifer M. S. Sucre, Bradley W. Richmond, Ana P. Serezani, Wyatt J. McDonnell, Simon B. Mallal, Matthew J. Bacchetta, James E. Loyd, Ciara M. Shaver, Lorraine B. Ware, Ross Bremner, Rajat Walia, Timothy S. Blackwell, Nicholas E. Banovich, Jonathan A. Kropski <<<

Single-cell RNA sequencing provides new insights into pathologic epithelial and mesenchymal remodeling in the human lung. <<<