Mingmeng Geng

Cela me rappelle un exposé www.ens.psl.eu/agenda/works... Plus précisément gauthierroussilhe.com

C'est un peu hors sujet, mais savoir un peu de latin et de grec, ça aide parfois à trouver un bon acronyme pour un modèle ou une méthode 😂

Many thanks again to my collaborators! Looking forward to meeting more people at upcoming conferences!

code_transformed: The Influence of Large Language Models on Code Coding remains one of the most fundamental modes of interaction between humans and machines. With the rapid advancement of Large Language Models (LLMs), code generation capabilities have begun to sign...

(--) Are Large Language Models Chameleons? An Attempt to Simulate Social Surveys arxiv.org/abs/2405.19323 (on another topic, oral presentation at ESRA)

And one new paper I‘ve mentioned to many people:
(5) code_transformed: The Influence of Large Language Models on Code arxiv.org/abs/2506.12014

(3) LLM as a Broken Telephone: Iterative Generation Distorts Information aclanthology.org/2025.acl-lon...
(4) Wikipedia in the Era of LLMs: Evolution and Risks arxiv.org/abs/2503.02879

(1) Human-LLM Coevolution: Evidence from Academic Writing aclanthology.org/2025.finding...
(2) The Impact of Large Language Models in Academia: from Writing to Speaking aclanthology.org/2025.finding...

Another unforgettable summer!
I was glad to present some of my recent work on "the impact of LLMs in society" at ACL (*3+1), IC2S2 (*2), ESRA, Youth in HD, and ICSSI.

Here are the papers and posters:

Wikipedia in the Era of LLMs: Evolution and Risks In this paper, we present a thorough analysis of the impact of Large Language Models (LLMs) on Wikipedia, examining the evolution of Wikipedia through existing data and using simulations to explore po...

hard to say arxiv.org/abs/2503.02879

Previous work:
(1) Is ChatGPT Transforming Academics' Writing Style arxiv.org/abs/2404.08627
(2) The Impact of Large Language Models in Academia: from Writing to Speaking arxiv.org/abs/2409.13686

[New preprint] Human-LLM Coevolution: Evidence from Academic Writing arxiv.org/abs/2502.09606

Hint 1: To delve or not to delve, that is the intricate question!
Hint 2: A short and easy-to-read paper!

Still the word frequency in arXiv abstracts! 👇👇👇

This translation has some issues 😂 ChatGPT does better. "Fly Over Southern University of Science and Technology: One Continuous Shot Covering 2,970 Mu in 11 Minutes · Double First-Class · SUSTech · Giant Campus · Aerial Campus View" PS: 1 mu ≈ 0.165 acres

飞越南方科技大学一镜到底2970亩11分钟·双一流·南科大·巨无霸·航拍校园_哔哩哔哩_bilibili -, 视频播放量 10343、弹幕量 35、点赞数 198、投硬币枚数 49、收藏人数 117、转发人数 80, 视频作者 旅拍者园子, 作者简介 所有视频素材均为个人原创，概不外借，也不得转载、录屏二创、或做其他任何用途。持续更新原创的一镜到底看名校、万米高空看中国等飞越系列。，相关视频：高考644上南科大不晓得亏不亏，南科大学生证的“含金量”，【毕业记录】逃离南科大，南科大23级硕士新生宿舍，...

bigger😉 www.bilibili.com/video/BV17J4...

Interesting, I know all four students in the video 😂 It was clearly filmed between 2013 and 2017, and the campus looks quite different now 🧐

You can find a little if you search her name 😬 My guess is that not many Chinese have moved here from X

Probably the only one that mentions her name so far:
bsky.app/profile/yili...

More discussion on X:
x.com/sunjiao123su...

Almost no discussion about Rosalind Picard here. Can I assume that most Chinese AI researchers are still on Twitter/X?

We are more interested in the density of LLM-style texts and its relative value (comparisons between categories and over time) than in establishing how many people are using LLMs – this can be estimated with the help of questionnaires, and it is not possible to get an accurate estimate only based on simulated data.

And, in an earlier paper 🧐😎 arxiv.org/abs/2404.08627

The Impact of Large Language Models in Academia: from Writing to Speaking Large language models (LLMs) are increasingly impacting human society, particularly in textual information. Based on more than 30,000 papers and 1,000 presentations from machine learning conferences, ...

Interesting work! In addition to what you mentioned, we also noticed that more LLM-style words have started to appear in the presentations of ML conferences: arxiv.org/abs/2409.13686

Maybe it's hard to define the previous distribution? (not PI, just intuition) 👀 "All happy families are alike; each unhappy family is unhappy in its own way."