September 28, 2025 – Kavli Institute Pre-Print Publications

Blocking apoptosis promotes survival and alters developmental dynamics of human retinal ganglion cells in retinal organoids

Posted by dbos September 28, 2025October 13, 2025Hopkins

Kavli Affiliate: Seth Blackshaw | Authors: Jingliang Simon Zhang, Brian Guy, Clayton P Santiago, Caterina Tiozzo, Meghana Sreenath, Ya-Wen Chen, Seth Blackshaw and Robert J Johnston, Jr | Summary: Retinal ganglion cells (RGCs) are the projection neurons that transmit visual information from the retina to the brain. In many species, a substantial proportion of RGCs […]

Continue..

Flexible Working Memory in the Peripheral Nervous System

Posted by dbos September 28, 2025October 13, 2025UCSD / Salk

Kavli Affiliate: Anastasia Kiyonaga | Authors: Sihan Yang, Yueying Dong and Anastasia Kiyonaga | Summary: Working memory (WM) representations that are distributed across the brain can be flexibly recruited to best guide behavior1–4. For instance, information may be represented relatively more strongly in visual cortex when a WM task requires fine visual detail, or more […]

Continue..

Search for Distant Hypervelocity Star Candidates Using RR Lyrae Stars

Posted by klaurent September 28, 2025Peking

Kavli Affiliate: Huawei Zhang | First 5 Authors: Haozhu Fu, Haozhu Fu, , , | Summary: Hypervelocity stars (HVSs) are stars with velocities exceeding their local escape velocities. Searching for HVSs and studying their origins can be an important way to study the properties of the Milky Way. In this paper, we utilize precise distances […]

Continue..

HunyuanImage 3.0 Technical Report

Posted by klaurent September 28, 2025Peking

Kavli Affiliate: Li Xin Li | First 5 Authors: Siyu Cao, Siyu Cao, , , | Summary: We present HunyuanImage 3.0, a native multimodal model that unifies multimodal understanding and generation within an autoregressive framework, with its image generation module publicly available. The achievement of HunyuanImage 3.0 relies on several key components, including meticulous data […]

Continue..

AutoPrune: Each Complexity Deserves a Pruning Policy

Posted by klaurent September 28, 2025Peking

Kavli Affiliate: Ke Wang | First 5 Authors: Hanshi Wang, Hanshi Wang, , , | Summary: The established redundancy in visual tokens within large vision-language models allows pruning to effectively reduce their substantial computational demands. Previous methods typically employ heuristic layer-specific pruning strategies where, although the number of tokens removed may differ across decoder layers, […]

Continue..

Probing Scalar-Mediated Sterile Neutrinos with Gravitational Wave and Colliders Signals

Posted by klaurent September 28, 2025Tokyo

Kavli Affiliate: Jia Liu | First 5 Authors: Qi Bi, Qi Bi, , , | Summary: We propose a UV-complete extension of the Standard Model in which a gauge-singlet scalar $S$ acquires a vacuum expectation value, generates a Majorana mass for a sterile neutrino $N$, and mixes with the Higgs field. This framework addresses neutrino […]

Continue..

RADAR: A Risk-Aware Dynamic Multi-Agent Framework for LLM Safety Evaluation via Role-Specialized Collaboration

Posted by klaurent September 28, 2025UCAS

Kavli Affiliate: Zheng Zhu | First 5 Authors: Xiuyuan Chen, Xiuyuan Chen, , , | Summary: Existing safety evaluation methods for large language models (LLMs) suffer from inherent limitations, including evaluator bias and detection failures arising from model homogeneity, which collectively undermine the robustness of risk evaluation processes. This paper seeks to re-examine the risk […]

Continue..

AdaPtis: Reducing Pipeline Bubbles with Adaptive Pipeline Parallelism on Heterogeneous Models

Posted by klaurent September 28, 2025Caltech

Kavli Affiliate: Wei Gao | First 5 Authors: Jihu Guo, Jihu Guo, , , | Summary: Pipeline parallelism is widely used to train large language models (LLMs). However, increasing heterogeneity in model architectures exacerbates pipeline bubbles, thereby reducing training efficiency. Existing approaches overlook the co-optimization of model partition, model placement, and workload scheduling, resulting in […]

Continue..

Determination of Aedes mosquito mating success by a rapidly-evolving female-controlled lock-and-key mechanism

Posted by dbos September 28, 2025October 7, 2025Rockefeller

Kavli Affiliate: Leslie B. Vosshall | Authors: Leah Houri-Zeevi, Madison M. Walker, Jacopo Razzauti, Anurag Sharma, H. Amalia Pasolli and Leslie B. Vosshall | Summary: Mosquitoes, the world’s deadliest animal, exemplify single-mating systems where females mate only once in their lifetime, making mate choice critically important for reproductive success and mosquito control. Despite this importance, […]

Continue..