1 code implementation • 6 Apr 2024 • Changjiang Gao, Hongda Hu, Peng Hu, Jiajun Chen, Jixing Li, ShuJian Huang
In this paper, we propose CLiKA, a systematic framework to assess the cross-lingual knowledge alignment of LLMs in the Performance, Consistency and Conductivity levels, and explored the effect of multilingual pretraining and instruction tuning on the degree of alignment.
no code implementations • 7 Mar 2024 • Changjiang Gao, Jixing Li, Jiajun Chen, ShuJian Huang
Drawing on the key-value memory interpretation of transformer feed-forward network blocks, we introduce the Composition Score, a novel model-based metric designed to quantify the degree of meaning composition during sentence comprehension.
1 code implementation • 29 Oct 2023 • Changjiang Gao, ShuJian Huang, Jixing Li, Jiajun Chen
Recent large language models (LLMs) have revealed strong abilities to understand natural language.
no code implementations • 18 Jan 2023 • Jixing Li, Xiaozhou Guo, Benzhe Dai, Guoliang Gong, Min Jin, Gang Chen, Wenyu Mao, Huaxiang Lu
This is always ignored and restricts the quantization performance.
no code implementations • COLING (CRAC) 2022 • Shulin Zhang, Jixing Li, John Hale
Pro-drop is commonly seen in many languages, but its discourse motivations have not been well characterized.
no code implementations • WS 2018 • Jixing Li, Murielle Fabre, Wen-Ming Luh, John Hale
The current study examined the role of syntactic structure during pronoun resolution.
no code implementations • WS 2018 • Jixing Li, Murielle Fabre, Wen-Ming Luh, John Hale
Typological differences between English and Chinese suggest stronger reliance on salience of the antecedent during pronoun resolution in Chinese.
no code implementations • WS 2016 • Jixing Li, Jonathan Brennan, Adam Mahar, John Hale
The relative contributions of meaning and form to sentence processing remains an outstanding issue across the language sciences.