[定位选校] 美国 NLP/LLM 方向 PhD 申请
18799
背景:
- 国内985 GPA 3.9; Upenn DS Master GPA 4.0
- 2封Upenn NLP教授research推荐信,1封PKU教授research推荐信
- NLP方向一作:1篇COLM,2篇AAAI;1篇ICML workshop;3篇在投(一作 ICLR/ICSAPP,二作ARR)
- 其他方向(e.g., IoT、HCI、Unlearning)二作以及之后挂名若干(8篇),大概率没用,只是来时的路
俩目前的兴趣:
1. To deepen our understanding of the internal mechanisms of LLMs and use these insights to improve their general capabilities. (可解释性+Inference adaptation/Training...)
2. To explore high-impact & previously unexplored problems, and adapt current models to better address these challenges. (Synthetic data/Post training/Code/Scientific Automation/AI4Sci...)
选校(2个以上match的教授):
JHU
CMU
UCSD
UMD
GT
NYU
Cornell
USC
UNC-Chapel Hill
UCSB
UW
match度不高的(1个match的):
UIUC
UCB
UW
UT-Austin
CU
1. 想问问大家我目前选校还要不要补充几个保底?(比如NCSU/NEU?)
一来我对目前的申请行情不太了解,二来考虑到funding问题...
2. 有没有其他专业方向强的学校补充(比如Harvard/Purdue/UMass/Rice/Duke?)
- 国内985 GPA 3.9; Upenn DS Master GPA 4.0
- 2封Upenn NLP教授research推荐信,1封PKU教授research推荐信
- NLP方向一作:1篇COLM,2篇AAAI;1篇ICML workshop;3篇在投(一作 ICLR/ICSAPP,二作ARR)
- 其他方向(e.g., IoT、HCI、Unlearning)二作以及之后挂名若干(8篇),大概率没用,只是来时的路
俩目前的兴趣:
1. To deepen our understanding of the internal mechanisms of LLMs and use these insights to improve their general capabilities. (可解释性+Inference adaptation/Training...)
2. To explore high-impact & previously unexplored problems, and adapt current models to better address these challenges. (Synthetic data/Post training/Code/Scientific Automation/AI4Sci...)
选校(2个以上match的教授):
JHU
CMU
UCSD
UMD
GT
NYU
Cornell
USC
UNC-Chapel Hill
UCSB
UW
match度不高的(1个match的):
UIUC
UCB
UW
UT-Austin
CU
1. 想问问大家我目前选校还要不要补充几个保底?(比如NCSU/NEU?)
一来我对目前的申请行情不太了解,二来考虑到funding问题...
2. 有没有其他专业方向强的学校补充(比如Harvard/Purdue/UMass/Rice/Duke?)
