LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging

Kavli Affiliate: Ke Wang | First 5 Authors: Ke Wang, Nikolaos Dimitriadis, Alessandro Favero, Guillermo Ortiz-Jimenez, Francois Fleuret | Summary: Large pre-trained models exhibit impressive zero-shot performance across diverse tasks, but fine-tuning often leads to catastrophic forgetting, where improvements on a target domain degrade generalization on other tasks. To address this challenge, we introduce LiNeS, […]


Continue.. LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging

LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging

Kavli Affiliate: Ke Wang | First 5 Authors: Ke Wang, Nikolaos Dimitriadis, Alessandro Favero, Guillermo Ortiz-Jimenez, Francois Fleuret | Summary: Fine-tuning pre-trained models has become the standard approach to endow them with specialized knowledge, but it poses fundamental challenges. In particular, textit{(i)} fine-tuning often leads to catastrophic forgetting, where improvements on a target domain degrade […]


Continue.. LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging

An Extreme Radio Fluctuation of Pulsar B1929$+$10

Kavli Affiliate: Renxin Xu | First 5 Authors: Zhengli Wang, Shunshun Cao, Jiguang Lu, Yulan Liu, Xun Shi | Summary: We report the detection of an extreme flux decrease accompanied by clear dispersion measure (DM) and rotation measure (RM) variations for pulsar B1929+10 during the 110-minute radio observation with the Five-hundred-meter Aperture Spherical radio Telescope […]


Continue.. An Extreme Radio Fluctuation of Pulsar B1929$+$10

Universality in the Near-Side Energy-Energy Correlator

Kavli Affiliate: Feng Yuan | First 5 Authors: Xiaohui Liu, Werner Vogelsang, Feng Yuan, Hua Xing Zhu, | Summary: We investigate the energy-energy correlator (EEC) of hadrons produced on the same side in $e^+e^-$ annihilation or in leading jets in $pp$ collisions. We observe a remarkable universality of the correlator. Using a non-perturbative transverse momentum […]


Continue.. Universality in the Near-Side Energy-Energy Correlator

Universality in the Near-Side Energy-Energy Correlator

Kavli Affiliate: Feng Yuan | First 5 Authors: Xiaohui Liu, Werner Vogelsang, Feng Yuan, Hua Xing Zhu, | Summary: We investigate the energy-energy correlator (EEC) of hadrons produced on the same side in $e^+e^-$ annihilation or in leading jets in $pp$ collisions. We observe a remarkable universality of the correlator. Using a non-perturbative transverse momentum […]


Continue.. Universality in the Near-Side Energy-Energy Correlator

Improving Parallel Program Performance with LLM Optimizers via Agent-System Interface

Kavli Affiliate: Ke Wang | First 5 Authors: Anjiang Wei, Allen Nie, Thiago S. F. X. Teixeira, Rohan Yadav, Wonchan Lee | Summary: Modern scientific discovery increasingly relies on high-performance computing for complex modeling and simulation. A key challenge in improving parallel program performance is efficiently mapping tasks to processors and data to memory, a […]


Continue.. Improving Parallel Program Performance with LLM Optimizers via Agent-System Interface

Improving Parallel Program Performance with LLM Optimizers via Agent-System Interface

Kavli Affiliate: Ke Wang | First 5 Authors: Anjiang Wei, Allen Nie, Thiago S. F. X. Teixeira, Rohan Yadav, Wonchan Lee | Summary: Modern scientific discovery increasingly relies on high-performance computing for complex modeling and simulation. A key challenge in improving parallel program performance is efficiently mapping tasks to processors and data to memory, a […]


Continue.. Improving Parallel Program Performance with LLM Optimizers via Agent-System Interface

ORCHID: A Chinese Debate Corpus for Target-Independent Stance Detection and Argumentative Dialogue Summarization

Kavli Affiliate: Ke Wang | First 5 Authors: Xiutian Zhao, Ke Wang, Wei Peng, , | Summary: Dialogue agents have been receiving increasing attention for years, and this trend has been further boosted by the recent progress of large language models (LLMs). Stance detection and dialogue summarization are two core tasks of dialogue agents in […]


Continue.. ORCHID: A Chinese Debate Corpus for Target-Independent Stance Detection and Argumentative Dialogue Summarization

Connection between Non-Axisymmetric Structures and Neutral Gas Distribution in Disk Galaxies

Kavli Affiliate: Luis C. Ho | First 5 Authors: Ze-Zhong Liang, Jing Wang, Hua Gao, Luis C. Ho, E. Athanassoula | Summary: Non-axisymmetric structures, such as bars and spiral arms, are known to concentrate molecular gas and star formation in galaxy centers, actively building up the pseudo-bulges. However, a direct link between the neutral (i.e., […]


Continue.. Connection between Non-Axisymmetric Structures and Neutral Gas Distribution in Disk Galaxies

Connection between Non-Axisymmetric Structures and Neutral Gas Distribution in Disk Galaxies

Kavli Affiliate: Jing Wang | First 5 Authors: Ze-Zhong Liang, Jing Wang, Hua Gao, Luis C. Ho, E. Athanassoula | Summary: Non-axisymmetric structures, such as bars and spiral arms, are known to concentrate molecular gas and star formation in galaxy centers, actively building up the pseudo-bulges. However, a direct link between the neutral (i.e., molecular […]


Continue.. Connection between Non-Axisymmetric Structures and Neutral Gas Distribution in Disk Galaxies