Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

How to pass the research AI interview? Reddit netizens Google interview experience sharing | attached resources

2025-01-19 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Internet Technology >

Share

Shulou(Shulou.com)06/02 Report--

2020-02-18 10:53:54

Thirteen from au Fei Temple

Qubit report | official account QbitAI

Most students will go through the process of internship before looking for a job, so the interview is essential.

A graduate student in the field of NLP at Reddit asked such a question:

What questions will be asked during the interview for a research internship? Do these questions mainly test your own experience or your level of understanding in the field?

Presumably this question also reflects the aspirations of most students.

So what kind of answer did Reddit netizens give?

Reddit netizens share interview experience

Netizen hyhieu, who participated in the first Google Brain Residents project, shared his interview experience in his comments.

First of all, most of the so-called "research interviews" means "tell me about your past research projects". I think the trick is to get the interviewer's attention because there is not enough time to do any real research during the interview process.

Be prepared to deal with some tough questions, for example, the interviewer may ask, "Why did you use XYZ in NuerIPS's 2019 paper? I think UVW will work better."

Second, the interview questions for "software engineering" are almost the same. Although it doesn't have much to do with AI, they keep asking these questions. All you have to do is go to LeetCode, or any website you like, and solve enough problems.

I don't think most researchers care about your performance in "software engineering".

Netizen mrpogiface, who recently attended an interview at Google and will join the Google brain team as an intern this summer, shared:

A total of three rounds of interviews were conducted, all on the phone.

For the first time, I talked about my research and papers, including why this field is important, what unsolved problems, what I have done, applications, and so on. This is the easiest round, because I can talk for hours, and I like my research. Then, I was asked some very basic machine learning questions, all about supervised learning.

The second interview looks more like a normal ML interview. They asked a rather complex NLP question. I discussed the most basic solutions for the current SOTA, more technical issues, but no code.

The third round is an ordinary Google-coded interview.

My background was math, but they didn't ask any "tricky" math brain teasers, nor did they ask any bizarre data questions. It's all very standard machine learning / deep learning stuff.

The preparatory work is very difficult. But if you are good at machine learning and have read the literature, it should be easy to pass the interview.

Netizen ajayrfhp also said:

It depends very much on the company itself. Generally will take the general machine learning, mathematics. The company doesn't want you to be an omnipotent expert in AI. Advanced interview questions will be tailored to your background and projects you have participated in.

Resource sharing

The quantum bit is also large.

Home has compiled a tutorial and resources related to AI, hoping to give you some help.

Columbia University Applied Machine Learning course 2020

Blog: http://www.cs.columbia.edu/~amueller/comsw4995s20/schedule/GitHub: https://github.com/amueller/COMS4995-s20 Video: https://www.bilibili.com/video/av86820508

Summary of the questions of "Jian Zhi Offer" & Summary of regular examination questions (Python implementation)

GitHub: https://github.com/shenweichen/coding_interviews

Free Chinese in-depth learning course, same as for undergraduates of Peking University

Course video: https://www.bilibili.com/video/av88056282/ course PPT: https://pan.baidu.com/s/1dN4NUtDs-tXLvGyKEO1MFQ extraction code: 126a

Book recommendation: "Natural language processing"

GitHub: https://github.com/jacobeisenstein/gt-nlp-class/blob/master/notes/eisenstein-nlp-notes.pdf

Is about to enter the spring recruitment stage, whether it is an internship or a formal position, are you ready?

Portal

Https://www.reddit.com/

-end-

Https://www.toutiao.com/i6794614355691307527/

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 0

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

Internet Technology

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report