#Tsinghua

[mysteriousquantumphysics]

Studying in China Remotely from Germany - Some Experiences

As I have mentioned in a previous entry, the past winter term I studied at Tsinghua University in Beijing. Unfortunately, I could not enter the country due to the Covid restrictions that have been still present when the semester started - nevertheless, I thought it might be valuable to share some experiences.

Studying Remotely

What had been a frustrating experience is that the exchange semester - which I had started organizing in summer 2021 - could not take place in person. The exchange semester started in September 22 and the information that exchange students cannot enter China was sent by the end of June 22. Thus, it was only roughly two months before the semester started until I knew for sure that I will not study on Tsinghua Campus. This had been unfortunate, because I was thinking about cancelling the exchange but it was too short to organize something else in Germany, as for instance an internship. I could have expected that it will not work anyway but somehow I kept some naive optimism until I knew for sure. Hence, after some considerations (Tsinghua expected a response already about one week after they sent the notice) I decided to do the exchange semester nevertheless. Even though this meant not having access to most of the experiences that make an exchange semester worthwhile and spending another semester mostly at home - even though there are nearly no Covid restrictions in Germany anymore. Back then, I was at least happy that I could avoid the risk of ending up in a harsh Chinese lockdown - the opposite happened: China gave up most of its Covid regulations. I’m happy and I hope that future exchange students will be more lucky than me in this respect.

Choice of Lectures

Since the lectures made up nearly everything of my exchange experience, it was also a little frustrating to see in the beginnning that the offered lectures in English are very limited at Tsinghua University - also regarding the point that one got access to the lecture lists only roughly a week before the semester started. In particular, nearly all useful lectures of the Physics Department were held in Chinese, which was unfortunate because I was enrolled as a student of this department and there were regulations that one was only allowed to do a very limited amount of credits outside the own department. Nevertheless, I tried to do the best out of it and attended at least one lecture (the only one which was in English and somehow useful for me) of the physics department about topological materials from an experimentalist’s standpoint. I already attended a theory course about this topic at TUM but at least I got a new perspective on some issues in this field.

Eventually, I also found interesting courses in the realm of computer science: one about theoretical informatics (automata theory) and machine learning. The latter was the most useful course in the whole semester because it covered a lot of different machine learning techniques, some of its mathematical background but it mostly focused on its application. The course offered a lot of programming exercises as well as a larger programming project which did not only help me to think through a more complex task but also gave me the opportunity to work together with Chinese students. In particular because I want to focus on numerical physics in my future and machine learning techniques become more and more prominent in physics, this computer science related lecture was very useful.

Last but not least, I also attended a course about Wittgenstein’s Tractatus, a lecture completely beyond my scientist-horizon. But it was a nice experience because it required a different kind of thinking than I am used to, even though analytical philosophy also covers aspects of philosophy of mathematics and science. I guess it is smart to learn a subject outside the bubble of quantitative science as well because it gives you some new perspectives which you usually easily ignore but shouldn’t.

Time Shift and Learning Mandarin

One further important point about studying remotely is the time shift of course. Between Munich and Beijing it is 6 or 7 hours difference (depending on the daylight saving time in Germany). Fortunately, most of the lectures were recorded anyway such that one could avoid living in another time zone. What one could not avoid was writing (midterm/final) exams in the middle of the night, what was definitely demanding.

Regarding learning Mandarin: I started learning Chinese one year before the exchange semester started, but to be honest I doubt it would have been enough for basic communcation in the supermarket. Even though Chinese has rather easy grammar, learning to understand the tones and the ambiguities in this language is the true challenge. However, it would have been truly useful to have more Mandarin skills: The online portals of the university are in Chinese of course and often I experienced that I could not access everything with the English counterparts. Another important point is that not all lecturers take English as instruction language so seriously: In one lecture many exercise materials had been in Chinese (and the topic was too technical to translate with baidu) such that exchange students had a clear disadvantage - but I want to emphasize that this depends on the lecturer - most of my lectures had been organized perfectly in English.

Conclusion and Hints

All in all, I hope I was able to make the best out of this exchange semester. I have mostly chosen lectures which were beyond my known horizon but nevertheless not useless in my discipline. Nevertheless, I could have attended lectures at my home universities which would have been much more favorable from an academic perspective. However, I think it was also nice to be forced to study outside the own discipline and to learn subjects I would have never decided to learn at TUM and LMU. Nonetheless, I still find it surprising that an elite university as Tsinghua offers most of its lectures in Chinese - in particular TUM and LMU are contrasting starkly regarding this, because here all lectures (at least at the physics departments) are held in English at the Masters level. 

Hence, I think this online exchange semester was a worthwhile experience despite all these issues. Finally, I am also happy that it is over and that I will start my Masters thesis soon here in Germany. I’m glad to have the chance to dive completely into (numerical) physics and related research questions! 

Finally, some hints if you’re planning to do an exchange semester in China (from my limited perspective as I haven’t been there in person)

When planning the exchange also develop a plan B: For me it would have made things so much easier if I had planned something in parallel (e.g. an internship in Germany) because then my decision of doing the exchange semester or not would not have been based on the lack of good alternatives.

Be patient: The administration (at least as I experienced it at Tsinghua) is rather slow and it happened that I never received answers to my e-mails. 

Be open-minded: As I explained in detail, lectures are mostly held in Chinese and one has to be flexible with the course choice in English. If you stick to the idea of remaining in your own discipline, an exchange semester will be a frustrating experience.

[chinemagazine]

Ouverture du deuxième Dialogue mondial de la jeunesse de Tsinghua

Le 2nd Dialogue mondial de la jeunesse de Tsinghua s'est tenu sur le thème "Harmonie et unité pour le plus grand bien commun"

Communiqué Tsinghua University – Le 29 août, le deuxième Dialogue mondial de la jeunesse de Tsinghua s’est tenu à l’université de Tsinghua. Le thème “Harmonie et unité pour le plus grand bien commun” a rassemblé plus de 100 représentants de la jeunesse de 35 pays et régions du monde. Ce rassemblement sert de creuset d’idées où de jeunes esprits dynamiques collaborent pour tracer des voies vers la…

[kbanews]

Safari Tsinghua

PUN ketika di Tsinghua. Buka puasanya di restoran Xinjiang. Dengan sate istimewanya itu. Di dekat kampus yang luasnya hampir 500 hektare ini. Di pinggir utara kota Beijing. “Boleh berapa orang?” tanya Lutfiya, mahasiswi S2 asal Lombok itu. “Berapa saja,” jawab saya. Rupanya dua kamar yang bisa digabung di resto itu hanya cukup untuk 20 orang. Maka hanya pendaftar pertama yang bisa gabung.…

[ammarthemystic]

Our Team was one of the finalists of this year’s U&AI | AI for SDGs Youth Bootcamp, hosted by the Institute for AI International Governance of Tsinghua University (I-AIIG) and supported by the United Nations Development Programme (UNDP). This was the culmination of a 5-month journey, aimed at encouraging youth from around the world to create innovative AI-based solutions to address real-world development challenges and help advance the Sustainable Development Goals (SDGs). This year, U&AI Camp had over 2,500 youth participants from 35 countries. Over the course of five months, these participants gained new knowledge and insights from 33 enlightening masterclasses delivered by renowned professionals from AI-related fields, and engaged in consistent online discussion to develop and fine-tune their projects. 14 teams made it to the bootcamp’s final stage, with AI solutions covering a wide range of topics including healthcare for vulnerable groups, reducing food waste, promoting gender equality, responding to climate change, protecting biodiversity, and promoting educational equality. #team #ai #healthcare #university #development #help #food #projects #sustainabledevelopment #climatechange #biodiversity #stage #gender #Tsinghua #catlaw #ailaw (at Tsinghua University) https://www.instagram.com/p/CnXMgZrrwOw/?igshid=NGJjMDIxMWI=

[cryptograndeenews]

In China, the dark side of robotization of production has been revealed: people work more, get less and are fired more often.

A panel of experts from the US, Canada and China analyzed China's workforce data as factory automation accelerates. It turned out that against the background of increased labor productivity, there were problems with employment, employment, income, and even with an increase in the birth rate.

Researchers from the University of Pittsburgh, Tsinghua University and the University of Toronto studied the data as part of China's five-year plan from 2016 to 2020 to increase factory automation. In these years, record amounts have been allocated for the robotization of industry in order to put China at the forefront of the new industrial revolution… Detail: https://bitcoingrandee.com/posts/34 NEWS

[jovialbasementbouquetblr]

2022: China Protests and Retribution

2022: China Protests and Retribution

This poem which appeared at Tsinghua University is circulating online in China. Many dots are added around the Chinese characters, apparently to frustrate the character recognition software that online censors used to scrub the Chinese networks clean. I found the two items below on on the Twitter feed at 冷山时评 @lengshanshipin Xi Jinping Step Down! Shared : this notice appeared at Tsinghua…

[pinkstarlightcomputer]

Sha Ching-Hwa Out of Dust 2019 Ink wash 76x97 cm.

[arcticboirealis]

Posting my Decapolice miiverse yeahs here

[acsisz]

Park vs Mall

These 4 has the Park background.

These 2 suddenly has the Mall background.

Idk if this is a Park, but they're all together at least. Why the change in the background shoots?

[decapolicemoment]

decapolice trailer

surprised that they confirmed he was top of his class (i knew he could do it). i wonder what carl placed

also going by their dynamic yk boston’s gonna be worrying about harvard later in the game (father figure of the year)

2. in most of the stuff we had before, the order was always harvard carl zhang mani mikey, though ig they log in the order harvard carl mani mikey zhang. does it mean anything? probably not

3. “My world is starting to move forward again”

☹️ therapy arc will go crazy

4.

two things. one, did they change where the bear comes in or is it going to be a reoccurring criminal? or does the bear stand for something else, maybe a criminal who they don’t have the identity of, so the bear’s a placeholder until they get better footage.

other thing is the lyric: “A line, once crossed. But compulsion’s strong.” He’s definitely going to go too far, or do some morally dubious things.

5.

i’m wondering if the hacker can take over characters data, in the sense that he can control them, cause it seems to be a Decasim employee who hands them the box/bomb/hack. Harvard had a gun pointed at him, so he was probably acting suspicious, and then the hack opened and changed stuff, like a trojan horse virus

6.

(i love her design) her hair reminds me of the clown’s, possibly related? she’s also the only one to have hair i wonder why

6.5, the lyrics at that moment “wings stained white as shadows descend without a trace” while showing her. in most cases (cause stained glass windows) things don’t get stained on purpose and they don’t stain themselves, usually it’s an outside force (staining a shirt, etc). perhaps she’s painted as a good person w/o her doing anything, while her true self (the clown) keeps doing things without anyone noticing. or maybe i’m reaching

7.

going back to 5, if the hacker’s able to override controls, then ig he got taken over or it’s js really convenient lighting. that or carl’s going to betray us/knows more than he should

8.

the lighting makes him look like he has eyebags go to sleep silly

9.

she might be human?? no name but atleast we have a full design

10.

the cat plush means smt im js wondering what. ❌’s for eyes, is harvard gonna almost die. maybe the clown has a lil harvard cat plush, for whatever reason he/she may need one

also “Just one thing remains”, his whole family is dead ig, but he only seems to care about his mother, probably because she was the only one murdered.

11. the thing that i still can’t understand

just about everything is confusing with this photo. why is his father ugly, and why does he have zero traits from him? why are his eyes GREEN?? neither of his parents have it so what’s up with that. why does he have the cat plush, and is there an in game reason why Harvard looks like it? who’s the cake for, it can’t be for him.

my biggest guess is that the hacker made it in his mother’s house in decasim, maybe as a way to mess with Harvard, show him what could have been or smt. maybe he’s trying to convince harvard to stay in decasim for whatever reason, showing his family alive and well, celebrating him becoming a detective, and the lil boy is his brother.

and finally 12.

who is telling him “we’ll meet again”? chances are it’s the clown, but maybe it’s his brother/clone/fakevard

anyways 10/10 trailer except for the 2026 release date

[materialsscienceandengineering]

New hydrogel can stretch to 15 times its original size

A team of molecular engineers at Tsinghua University, in China, has developed a new type of hydrogel that can stretch to 15 times its original size and then snap back to its original form. In their study, published in the journal Science, the group modified the process normally used to create hydrogels to produce a new one that is far more elastic. Hydrogels are known for their stretchiness—they can be pulled like taffy or a rubber band. But most do not snap back to their original form very well, making them stretchy but not elastic. Additionally, they can only be stretched in one direction. Currently, hydrogels are generally made by creating compounds with cross-linked polymers linked by water molecules. In this new effort, the team in China sought to improve the characteristics of a hydrogel by making changes to the fabrication method.

Read more.

[hasdrubal-gisco]

male loneliness epidemic isn't real because you can make a tinder, use any picture of you with a lanyard, set your job to "team lead at [google your local aeronautics/arms/miltech company]" and a beautiful 5'6" (tall !) chinese shawty will be eager to meet and marry you

[call-me-rucy]

So I know fuck all about video edition or hard subbing or Youtube, but I tried to share what I spent today doing: a translation of the Decapolice demo.

I hope you can forgive the terrible quality and have fun watching Harvard and co. solve the robbery.

[mysteriousquantumphysics]

Data Science meets the Many Body Problem

Since the machine learning course I did this semester at Tsinghua university was mainly focused on typical data science applications, I was curious in how far those methods can be applied in physics. Of course it is nothing new that neural networks can in principle also be used for physical applications - however, tensor network methods still seem to be dominant in the field of numerical many body physics. Thus, I decided to dive in a little into the literature about the usage of Restrictive Boltzmann Machines (RBM) in many body physics.

What are RBMs?

Usually, RBMs are used for instance for recommendation tasks (e.g. video recommendations on video platforms) and many more. In general, it is a unsupervised learning technique which makes use of minimizing its "energy". Thus, the intuition behind RBMs is, despite their data science applications, already related to physics: We will see that it is no surprise that "Boltzmann" is part of the name of this method. An RBM utilizes input data, tries to extract meaningful features from it and wants to find the probability distribution over the input. This follows the physical intuition as follows: The RBM is a neural network with two layers, a hidden and a visible layer, where each node can adopt binary values. An example network looks like:

where the x denote the visible nodes, the h the hidden nodes and W denotes the weights between both layers. Note that there are no links in between the nodes of a single layer; this is why these networks are called restricted Boltzmann machines.

The network is governed by a corresponding energy function as:

where we also have the offsets of the single nodes (a for the visible nodes and b for the hidden nodes). Given this energy function, one can determine the probability distribution by the Boltzmann distribution where Z is a partition function, as familiar from classical statistical physics. As usual, the energy is supposed to be minimized, what is done by the learning algorithm of the RBM but this should not explained here since it would go beyond the scope of a brief blog entry. At this point I'd only like to mention that there are some difficulties determining e.g. the partition function (which is intractable in general) and that this requires some sophisticated algortihms. If you're interested in this and how the RBMs work exactly, a neat and far more rigorous introduction into RBMs can be found here.

One side note at this point: RBMs were introduced by Geoffrey Hinton after John Hopfield (a physicist) invented the so-called Hopfield networks which are also such an energy based mechanism, based on the physical intuition of Ising models.

Note that so far we only talked about the RBMs as they are used in data science - despite their physical intuition, they had so far nothing to do with neither quantum mechanics nor the many body problem. This is what comes next.

How can this be linked to condensed matter?

As introduced in [1], an RBM that can represent a quantum many body state would look like this:

In comparison to the previous network we changed the labels from x to σ, where the σ's denote e.g. spin 1/2 configurations, bosonic occupation numbers and so on. For them one has to choose a basis, e.g. the σ^z basis. This configuration can be summarized in the set S. Hence, the visible nodes are the N physical nodes of the system. The M hidden nodes h play the role of auxiliary (spin) variables. The authors describe the understanding of such a neural-network quantum state as follows: "The many-body wave function is a mapping of the N−dimensional set S to (exponentially many) complex numbers which fully specify the amplitude and the phase of the quantum state. The point of view we take here is to interpret the wave function as a computational black box which, given an input manybody configuration S, returns a phase and an amplitude according to Ψ(S)" [1, p.2]. Thus, one gives a certain spin configuration as input and the RBM generates the state, in the following form:

Thus, once can recognize that the form of such a neural-network quantum state adopts a similar form as the aforementioned Boltzmann distribution (exponential of energy function). However, there is additionally a sum over all possible hidden configurations which specifies the full state. After setting a state up in this form, one aim could be to find the ground state corresponding to a certain Hamiltonian, and according to the authors of [1], their RBM method gives decent results for this task!

Similarity to tensor networks

Interesting is, that this framework (even though it appears very different) has some analogous quantities as tensor network states. For example, the representational quality of a neural-network quantum state can be increased by increasing the number of hidden states: Thus, the ratio M/N plays a similar role as the bond dimension of a tensor product state! There are many more similarities, which should not be discussed here but can be found in [3].

Nevertheless, I'd like to mention an important distinction, which is also crucial for tensor network states, because some algorithms (DMRG etc.) can only handle area law states properly. While volume law states have an entanglement entropy which scales with the volume of the partitions of a state, an area law state has a scaling only proportionally with the area of the cut. Area law states can be handled better numerically, because the bond dimension of tensor network states explode for volume law states (more on tensor networks and area law can be found in [4]). According to [2] the difference between area law states and volume law states can be captured in a neat way with RBM states: While volume law states must have full connections between the hidden and physical nodes, an area law state has fewer links - this imposes locality in a sense. The RBM states thus give a neat intuition between the differenece of both kinds of states.

All in all, RBMs seem to be an interesting approach to connect both data science methods and many body physics. It may be that they have strengths which the usual tensor networks approaches lack: for instance, the authors of [2, p.888] claim that it might be possible that RBMs might be able to handle volume law states better than usual tensor network approaches do, which would be of course a major benefit. Since I haven't heard of this approach within the condensed matter framework before, I'm very curious how the importance of this method will evolve in future research!

--- References:

[1] Carleo, Troyer, Solving the Quantum Many-Body Problem with Artificial Neural Networks, arXiv:1606.02318

[2] Melco, Carleo, Carrasquilla, Cirac, Restricted Boltzmann machines in quantum physics, https://doi.org/10.1038/s41567-019-0545-1

[3] Chen, Cheng, Xie, Wang, Xiang, Equivalence of restricted Boltzmann machines and tensor network states, arXiv:1701.04831

[4] Hauschild, Pollmann, Efficient numerical simulations with Tensor Networks: Tensor Network Python (TeNPy), arXiv:1805.00055

[megamidevice]

Hatsune Miku (Mech Mirai 2020.5 Motor ver.) • Vocaloid 1/7 Scale Doujin Figure by The Future Animation x Game Student Technology Team Tsinghua University *with Bluetooth speaker function

[jovialbasementbouquetblr]

2014: Tsinghua U Public Emergency Response Plan

2014: Tsinghua U Public Emergency Response Plan

With the outbreak of student protests across China about Covid lockdowns and other grievances, many Chinese universities must be dusting off their emergency plans. The 2014 Tsinghua University plan translated below is one such plan. Some of these plans address emergencies that college deans the world over must worry about. Other address how to calm down and disperse unruly student protesters…