Yukang Xie | Searcharxiv

arXiv2026

Testing the Distance Duality Relation with Cosmological Observations at high Redshift using Artificial Neural Network

The cosmic Distance Duality Relation (DDR) is a fundamental prediction of metric gravity under photon number conservation. In this work, we perform a model-independent test of the DDR using Pantheon+ type Ia supernovae (SN Ia), \emph{Fermi} gamma-ray bursts (GRBs) with the FULL and GOLD samples, the Dark Energy Spectroscopic Instrument (DESI) Data Release 2 (DR2) baryon acoustic oscillation (BAO) measurements, and the galaxy-scale strong gravitational lensing (SGL) system samples at high redshift $0.01 < z \lesssim 8$ using an artificial neural network (ANN) approach. Our results show that the standard DDR is consistent with cosmological observations at high redshift within the $\sim 2 σ$ confidence level.

astro-ph.CO↗

arXiv2023

Making Small Language Models Better Multi-task Learners with Mixture-of-Task-Adapters

Recently, Large Language Models (LLMs) have achieved amazing zero-shot learning performance over a variety of Natural Language Processing (NLP) tasks, especially for text generative tasks. Yet, the large size of LLMs often leads to the high computational cost of model training and online deployment. In our work, we present ALTER, a system that effectively builds the multi-tAsk Learners with mixTure-of-task-adaptERs upon small language models (with <1B parameters) to address multiple NLP tasks simultaneously, capturing the commonalities and differences between tasks, in order to support domain-specific applications. Specifically, in ALTER, we propose the Mixture-of-Task-Adapters (MTA) module as an extension to the transformer architecture for the underlying model to capture the intra-task and inter-task knowledge. A two-stage training method is further proposed to optimize the collaboration between adapters at a small computational cost. Experimental results over a mixture of NLP tasks show that our proposed MTA architecture and the two-stage training method achieve good performance. Based on ALTER, we have also produced MTA-equipped language models for various domains.

cs.CL↗