Researcher @UZH_en

Joined December 2019
70 Photos and videos
Pinned Tweet
3 Jul 2024
Should I use Macro F1 or Accuracy? Why not Kappa? Why do some use this, and others that? What's actually evaluated here? 😵‍💫 Happy to share the final version of this paper on multi-class classification evaluation: direct.mit.edu/tacl/article/… #machinelearning #nlproc #ml
6
21
2,189
juri retweeted
There's a lot of research papers on AI for science focused on ML engineering, but as of yet I haven't seen any examples of an AI-discovered ML algorithm that has actually seen mainstream acceptance. Are there any examples of this?
5
2
101
14,706
juri retweeted
as an AC: LLM assisted reviews are terrible. LLM assisted author responses are terrible. Even if the authors and reviewers do take the effort to carefully guide the LLMs so they reflect their points precisely (and not all do) the result is a wall of text that is just painful.
10
5
98
12,134
juri retweeted
Code volume does not represent productivity.
Massive output uptick due to agentic AI. Complete flat adoption.
98
68
712
65,919
juri retweeted
It’s really incredible the absolute AI GARBAGE that people are comfortable sending to their coworkers and bosses There’s a good chance productivity will actually *decrease* as AI adoption increases because everyone is busy wading through AI slop
138
147
3,256
174,445
juri retweeted
It's so cringe when real people I otherwise know and respect post obvious AI slop on social media, particularly when they're (supposedly) expressing their feelings. Authenticity is so rare and valuable these days, and it's sad to see people just cede it from the get-go
4
11
120
28,412
May 18
Re LLMs as reviewers to cope with submission load. LLMs and AI models have essentially been trained on a snapshot of the past, afaik with a gap of up to 2-3 years or even more until now. How can they be good reviewers in peer-review, and on what metric?
3
541
juri retweeted
Wow, so much whining about arXiv’s steps to reduce AI slop. So easy to deal with for authors who actually read their own papers before submitting them.
7
19
276
6,434
The backlash against arXiv is a bit odd. All they're asking is that you read your papers before submitting them.
44
142
2,134
80,911
juri retweeted
One of the biggest problems with using LLMs as a google replacement for programming, is that getting zero relevant results on google used to be a signal that you had the wrong idea about the root cause. Whereas LLMs will happily indulge any terrible idea you suggest.
141
613
10,123
195,766
juri retweeted
119
1,528
33,255
511,960
juri retweeted
📢 Postdoc Position in NLP @ UTN in Nuremberg, Germany I am looking for a full-time postdoctoral researcher (A13/E13, initial contract for 3 yrs) starting July 2026 or as soon as possible thereafter. Focus on implicit & underspecified language, background knowledge and/or biases.
1
7
35
4,372
juri retweeted
looks like there's gonna be around 40k neurips submissions? the biggest exponential in ai right now is slop
15
8
274
24,659
May 4
Just raised some points in the rebuttal as reviewer, after thoughtful author responses! Sadly our own paper doesn't seem to receive the honor, even though reviewers thanked us for having added an experiment and issues "clarified" or "resolved" 🥲
2
93
NLP relies on Linguistics & that's capital RELIES! It's RELIES, which encapsulates six major facets where linguistics contributes to NLP: Resources, Evaluation, Low-resource settings, Interpretability, Explanation, and the Study of language. Read it at: doi.org/10.1162/coli_a_00560
6
22
1,044
Apr 21
Deadline in 2 days! Last chance to register for our novel shared task at CLEF 2026 - Classifying the relations of locations and persons that are mentioned in a news text! #nlproc #machinelearning #datascience hipe-eval.github.io/HIPE-202…
56
Apr 17
I see more and more papers exploring similar concepts that were already explored in the BERT and other previous eras. This is cool, but then it's kind of disappointing if the related work just dates max 2-3 years back and basically ignores all this.
2
10
597
Apr 13
I see more papers that in their bibliography mention hundreds of names for a single paper often using more than 1 page. I think instead it makes sense to use this APA recommendation: A, B, C, ... D (2019), where C is the 19-th name and D the last author. apastyle.apa.org/blog/more-t…
125
Because students now use ChatGPT and other AI tools to outline essays, skim readings, and solve homework problems—to perform nearly every assigned task—the majority of the writing they encounter will be AI-generated.
15
71
186
21,356
Mar 24
Accepted at EACL 2026 System Demonstrations 😊: Github: github.com/flipz357/XPLAINSI… Paper: aclanthology.org/2026.eacl-d… #nlproc #machinelearning #eacl2026
11
400