Kavli Affiliate: Ke Wang | First 5 Authors: Zimu Lu, Zimu Lu, , , | Summary: Agent systems powered by large language models (LLMs) have demonstrated impressive performance on repository-level code-generation tasks. However, for tasks such as website codebase generation, which depend heavily on visual effects and user-interaction feedback, current code agents rely only on […]

Kavli Affiliate: Jing Wang | First 5 Authors: Hua Yuan, Hua Yuan, , , | Summary: Parameter transfer is a central paradigm in transfer learning, enabling knowledge reuse across tasks and domains by sharing model parameters between upstream and downstream models. However, when only a subset of parameters from the upstream model is transferred to […]

Kavli Affiliate: Jing Wang | First 5 Authors: Sicong Leng, Sicong Leng, , , | Summary: Large multimodal reasoning models have achieved rapid progress, but their advancement is constrained by two major limitations: the absence of open, large-scale, high-quality long chain-of-thought (CoT) data, and the instability of reinforcement learning (RL) algorithms in post-training. Group Relative […]

Kavli Affiliate: Luis C. Ho | First 5 Authors: John D. Silverman, John D. Silverman, , , | Summary: Contamination-free assessments of the interstellar medium and star formation in quasar host galaxies, particularly based on the far-infrared, offer insights into the role of supermassive black holes in galaxy evolution. Motivated by predictions of quasar heating […]

Kavli Affiliate: Lile Wang | First 5 Authors: Shunyuan Mao, Shunyuan Mao, , , | Summary: Accretion disks are ubiquitous in astrophysics, appearing in diverse environments from planet-forming systems to X-ray binaries and active galactic nuclei. Traditionally, modeling their dynamics requires computationally intensive (magneto)hydrodynamic simulations. Recently, Physics-Informed Neural Networks (PINNs) have emerged as a promising […]

Kavli Affiliate: Zhuo Li | First 5 Authors: Huizhen Shu, Huizhen Shu, , , | Summary: Achieving robust safety alignment in large language models (LLMs) while preserving their utility remains a fundamental challenge. Existing approaches often struggle to balance comprehensive safety with fine-grained controllability at the representation level. We introduce LATENTGUARD, a novel three-stage framework […]

Kavli Affiliate: Gregory J. Herczeg | First 5 Authors: Min Fang, Min Fang, , , | Summary: We re-evaluate the star formation history of the nearby Sco-Cen OB Association with a comprehensive analysis of Gaia XP spectra of more than 7,800 potential members. New spectral classifications are obtained by fitting individual XP spectra with templates […]

Kavli Affiliate: Kohei Inayoshi | First 5 Authors: Kohei Inayoshi, Kohei Inayoshi, , , | Summary: The birth of seeds of massive black holes (BHs) and nascent galaxies at cosmic dawn takes place in dense gaseous environments, which play a crucial role in shaping their coevolution and radiation spectra. We investigate gas accretion during the […]

Kavli Affiliate: Claudio Ricci | First 5 Authors: Marco Troncoso, Marco Troncoso, , , | Summary: Galaxy mergers are believed to play an important role in triggering rapid supermassive black hole (SMBH) growth. As merging nuclei approach each other, the physical properties of the participating galaxies and the associated SMBH growth are expected to evolve […]

Kavli Affiliate: Xuebing Wu | First 5 Authors: Di Wu, Di Wu, , , | Summary: Wide-separation lensed quasars (WSLQs) are a rare subclass of strongly lensed quasars produced by massive galaxy clusters. They provide valuable probes of dark-matter halos and quasar host galaxies. However, only about ten WSLQ systems are currently known, which limits […]