​​​​​Oxytocin's Selective Facilitation of Socially Reinforced Learning 

Previous studies have found that given social feedbacks, learners’ procedural learningcan be reinforced accordingly. This study aims to find out whether this reinforcement can be further enhanced or reversely reduced by oxytocin, and which brain areas & neural networks are involved in this process. ​​

RALT (Reinforced Association Learning Test) Paradigm :

​​​​Several 3-digit numbers are randomly assigned to two categories: A or B. Given a random number, subjects are asked to assign the number to a certain category, and will be given a feedback (social, semi-social or non-social) indicating whether they are right. Test for the same number set will be repeated for multiple times, allowing subjects to gradually develop knowledge of the initial categorization through reinforced learning.   

Publication: ​​
[1] Hu J.*, Song Q.*, Becker B.*, Luo L., Gao S., Gong Q., Hurlemann R. and Kendrick K.M. (2015). Oxytocin selectively facilitates learning with social feedback and increases activity and functional connectivity in emotional memory and reward processing regions, Human Brain Mapping, doi: 10.1002/hbm.22760.​​

*These authors contributed equally to this work​​


Chinese-English Bilinguals' Lexical Access​​​

How bilinguals’ mental lexicon is represented and retrieved has long been a hot topic in the field of bilingualism research. Large numbers of researches have proven that the bilingual lexicon representation is divided into two dimensions: lexical level and conceptual level. While the two languages’ lexical representations are independent, they share a common conceptual representation. (Zeelenberg & Pecher 2003; De Groot, 2005;Jared & Kroll, 2001; Li Li et al.,2008). However, how’s the two independent lexical representations interact with the common conceptual representation remains controversial. ​

Long-term Repetition Priming:
The whole experiment is composed of two sections, a encoding section (study section) and a retrieving section (test section). By adjusting the relationship between materials in the two sections, we can observe the influence of “whether the material has been previously learned in the study section” on participants’ responses to the material in the test section.​​​

Song Qi and Jiehui Hu. Orthographic Lexical Access is L2-Proficiency-Dependent in
Different-Script Bilinguals: Evidence from Long-Term Repetition Priming. Foreign Lan-guage Annals. (Submitted)​​​


Collective Electrotaxis of Epithelial Cell

While migration behavior under electric fields for single cells (Ex. Human immune T cells) has been thoroughly studied, the electrotaxis mechanism for epithelial cells(Ex. MDCK cells) remains largely unknown. The present study seeks to address the issue using a combination of numerical modeling/simulation and electrotaxis experiments.​


Modeling Earth Air Health via Network Dynamics​​​​​​​​​​​​​​​​​​​​
(ICM 2013 Contest Paper with Honorable Mention)​

We construct a directed local temporal network with ​each node representing a country. Parameters associated with a node describe how air pollution is generated, resolved, and distributed. Links among the nodes represent possible pathways for pollution transfer. We then model a global node with bifurcation dynamic and connect it to the local network. Local forcings’ influence ​on the global forcing and feedbacks from the global node are simulated ​by correlating the entropy of the local network and the chaos index in the global bifurcation. We set 2 measures at the local level and global level, reflected by local pollution accumulation and global bifurcation status respectively. The model parameters are estimated via analytical hierarchy process.​