Florian Eichin
@florian-eichin.com
PhD candidate at LMU Munich. Representations, model and data attribution, training dynamics.
Strong opinions on coffee and tea ☕
https://florian-eichin.com
Strong opinions on coffee and tea ☕
https://florian-eichin.com
Reposted by Florian Eichin
Thank you to our great Munich Center for Machine Learning (@munichcenterml.bsky.social) for featuring me in this research film! Lots of great films clips with my MCML colleagues are available on MCML's YouTube channel. #ai #aiethics #mcml #philosophy youtu.be/KUqiY8o1yng?...
August 7, 2025 at 9:19 AM
Thank you to our great Munich Center for Machine Learning (@munichcenterml.bsky.social) for featuring me in this research film! Lots of great films clips with my MCML colleagues are available on MCML's YouTube channel. #ai #aiethics #mcml #philosophy youtu.be/KUqiY8o1yng?...
Reposted by Florian Eichin
🕺🏼swing by our poster in Hall 4/5 on Wednesday, July 30 at 11:00 to chat with @florian-eichin.com and I to find out the answers to these questions
🛎️ bonus: to see the full poster 🫣🧩
#ACL2025 #NLProc
🛎️ bonus: to see the full poster 🫣🧩
#ACL2025 #NLProc
July 23, 2025 at 2:03 PM
🕺🏼swing by our poster in Hall 4/5 on Wednesday, July 30 at 11:00 to chat with @florian-eichin.com and I to find out the answers to these questions
🛎️ bonus: to see the full poster 🫣🧩
#ACL2025 #NLProc
🛎️ bonus: to see the full poster 🫣🧩
#ACL2025 #NLProc
Reposted by Florian Eichin
📝Probing LLMs for Multilingual Discourse Generalization Through a Unified Label Set
🔎Do LLMs encode and generalize discourse knowledge across languages?
👥 @florian-eichin.com @janetlauyeung.bsky.social @mhedderich.bsky.social @barbaraplank.bsky.social
🔗 arxiv.org/abs/2503.10515
📁Main - Long
🔎Do LLMs encode and generalize discourse knowledge across languages?
👥 @florian-eichin.com @janetlauyeung.bsky.social @mhedderich.bsky.social @barbaraplank.bsky.social
🔗 arxiv.org/abs/2503.10515
📁Main - Long
July 23, 2025 at 12:30 PM
📝Probing LLMs for Multilingual Discourse Generalization Through a Unified Label Set
🔎Do LLMs encode and generalize discourse knowledge across languages?
👥 @florian-eichin.com @janetlauyeung.bsky.social @mhedderich.bsky.social @barbaraplank.bsky.social
🔗 arxiv.org/abs/2503.10515
📁Main - Long
🔎Do LLMs encode and generalize discourse knowledge across languages?
👥 @florian-eichin.com @janetlauyeung.bsky.social @mhedderich.bsky.social @barbaraplank.bsky.social
🔗 arxiv.org/abs/2503.10515
📁Main - Long
Some recommendations for #ACL2025 👇
(join me and @janetlauyeung.bsky.social to talk about discourse generalization and probing!)
(join me and @janetlauyeung.bsky.social to talk about discourse generalization and probing!)
Headed to ACL? MaiNLP & our most recent work will be there too👥📄
Come see what we’ve been working on!
Come see what we’ve been working on!
July 23, 2025 at 12:38 PM
Some recommendations for #ACL2025 👇
(join me and @janetlauyeung.bsky.social to talk about discourse generalization and probing!)
(join me and @janetlauyeung.bsky.social to talk about discourse generalization and probing!)
Reposted by Florian Eichin
Headed to ACL? MaiNLP & our most recent work will be there too👥📄
Come see what we’ve been working on!
Come see what we’ve been working on!
July 23, 2025 at 12:30 PM
Headed to ACL? MaiNLP & our most recent work will be there too👥📄
Come see what we’ve been working on!
Come see what we’ve been working on!
Reposted by Florian Eichin
XAI’s dogwater performance on the 2025 IMO confirms that their Grok 4 benchmark claims were hot air. Their eye popping metrics were down to the following innovations:
- train on test
- train on test
- train on test
- train on test
- train on test
- train on test
July 20, 2025 at 5:15 AM
XAI’s dogwater performance on the 2025 IMO confirms that their Grok 4 benchmark claims were hot air. Their eye popping metrics were down to the following innovations:
- train on test
- train on test
- train on test
- train on test
- train on test
- train on test
Reposted by Florian Eichin
📄 [ACL 2025 main] LLMs instead of Human Judges? A Large Scale Empirical Study across 20 NLP Evaluation Tasks (doi.org/10.48550/arX...)
LLMs instead of Human Judges? A Large Scale Empirical Study across 20 NLP Evaluation Tasks
There is an increasing trend towards evaluating NLP models with LLMs instead of human judgments, raising questions about the validity of these evaluations, as well as their reproducibility in the case...
doi.org
July 18, 2025 at 10:19 AM
📄 [ACL 2025 main] LLMs instead of Human Judges? A Large Scale Empirical Study across 20 NLP Evaluation Tasks (doi.org/10.48550/arX...)
Reposted by Florian Eichin
📄 [ACL 2025 main] Circuit compositions: Exploring Modular Structures in Transformer-Based Language Models (doi.org/10.48550/arX...)
Circuit Compositions: Exploring Modular Structures in Transformer-Based Language Models
A fundamental question in interpretability research is to what extent neural networks, particularly language models, implement reusable functions through subnetworks that can be composed to perform mo...
doi.org
July 18, 2025 at 10:19 AM
📄 [ACL 2025 main] Circuit compositions: Exploring Modular Structures in Transformer-Based Language Models (doi.org/10.48550/arX...)
Reposted by Florian Eichin
Interpretability meets Discourse. Congratulations to
@florian-eichin.com to his first ACL paper 🎉
@florian-eichin.com to his first ACL paper 🎉
🦙 how well do LLMs encode discourse knowledge? does that generalize across languages?
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
July 10, 2025 at 1:19 PM
Interpretability meets Discourse. Congratulations to
@florian-eichin.com to his first ACL paper 🎉
@florian-eichin.com to his first ACL paper 🎉
Paper alert 🛎️
🦙 how well do LLMs encode discourse knowledge? does that generalize across languages?
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
July 10, 2025 at 12:40 PM
Paper alert 🛎️
Reposted by Florian Eichin
🦙 how well do LLMs encode discourse knowledge? does that generalize across languages?
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
July 10, 2025 at 12:38 PM
🦙 how well do LLMs encode discourse knowledge? does that generalize across languages?
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
🛎️ in our #ACL2025 paper, we uncover fascinating trends about multilingual discourse representations!
joint work w/ @florian-eichin.com @barbaraplank.bsky.social @mhedderich.bsky.social
📄 arxiv.org/abs/2503.10515
Reposted by Florian Eichin
I’ll be at @icmlconf.bsky.social next week presenting NoLiMa!
Poster on Tue July 15, 4:30–7pm (E-2312).
Happy to grab a coffee and chat about long-context, memory, research, or just to catch up.
I’ll be in Toronto for a couple of days after the conference, let me know if you’re around!
Poster on Tue July 15, 4:30–7pm (E-2312).
Happy to grab a coffee and chat about long-context, memory, research, or just to catch up.
I’ll be in Toronto for a couple of days after the conference, let me know if you’re around!
July 9, 2025 at 1:53 PM
I’ll be at @icmlconf.bsky.social next week presenting NoLiMa!
Poster on Tue July 15, 4:30–7pm (E-2312).
Happy to grab a coffee and chat about long-context, memory, research, or just to catch up.
I’ll be in Toronto for a couple of days after the conference, let me know if you’re around!
Poster on Tue July 15, 4:30–7pm (E-2312).
Happy to grab a coffee and chat about long-context, memory, research, or just to catch up.
I’ll be in Toronto for a couple of days after the conference, let me know if you’re around!
Reposted by Florian Eichin
Caught some great moments at #MCML Munich AI Day 2025 last week📍
From sharp keynotes to poster debates. Our team had the chance to show some recent work, join the conversations, and bring back plenty of food for thought🧠🗣️📊
From sharp keynotes to poster debates. Our team had the chance to show some recent work, join the conversations, and bring back plenty of food for thought🧠🗣️📊
Last week, #MCML Munich AI Day 2025 kicked off with keynotes by Julia Schnabel and Tina Eliassi-Rad, brilliantly moderated by Eva Schulz.
July 9, 2025 at 8:14 AM
Caught some great moments at #MCML Munich AI Day 2025 last week📍
From sharp keynotes to poster debates. Our team had the chance to show some recent work, join the conversations, and bring back plenty of food for thought🧠🗣️📊
From sharp keynotes to poster debates. Our team had the chance to show some recent work, join the conversations, and bring back plenty of food for thought🧠🗣️📊
Reposted by Florian Eichin
The study is here but gated: journals.sagepub.com/doi/10.3102/...
I’d be curious how these dynamics play out in our NLP review crisis. My hunch: many conscientious volunteers might be junior women. That time comes at a cost; chasing slackers means less time for rebutting my own reviews.
I’d be curious how these dynamics play out in our NLP review crisis. My hunch: many conscientious volunteers might be junior women. That time comes at a cost; chasing slackers means less time for rebutting my own reviews.
In a study of professors, women got 378 new work requests over 4 weeks vs 118 for men. Women spent more time on service, advising & teaching; men on research. Orgs should track who is taking extra duties & ensure they are rewarded and distributed fairly. www.forbes.com/sites/kimels...
Being Too Helpful At Work Can Hurt Your Career—Here’s How To Say No
Women are more likely to take on behind-the-scenes duties at work—extra tasks like onboarding or event planning—and it's hurting their careers. Here's how to say no.
www.forbes.com
July 8, 2025 at 5:56 AM
The study is here but gated: journals.sagepub.com/doi/10.3102/...
I’d be curious how these dynamics play out in our NLP review crisis. My hunch: many conscientious volunteers might be junior women. That time comes at a cost; chasing slackers means less time for rebutting my own reviews.
I’d be curious how these dynamics play out in our NLP review crisis. My hunch: many conscientious volunteers might be junior women. That time comes at a cost; chasing slackers means less time for rebutting my own reviews.
Reposted by Florian Eichin
Thanks for the invitation to the Freiburg Institute for Advanced Studies (FRIAS) to give this year's Hermann-Paul-Center Lecture lnkd.in/d_wUeDfY
I enjoyed the visit, the great audience, and the stay in this lovely city.
Thank you
#blackforest #freiburg #breisgau
I enjoyed the visit, the great audience, and the stay in this lovely city.
Thank you
#blackforest #freiburg #breisgau
July 4, 2025 at 11:00 AM
Thanks for the invitation to the Freiburg Institute for Advanced Studies (FRIAS) to give this year's Hermann-Paul-Center Lecture lnkd.in/d_wUeDfY
I enjoyed the visit, the great audience, and the stay in this lovely city.
Thank you
#blackforest #freiburg #breisgau
I enjoyed the visit, the great audience, and the stay in this lovely city.
Thank you
#blackforest #freiburg #breisgau
Reposted by Florian Eichin
Here's the preprint---basically, you can just use concept erasure (github.com/EleutherAI/c...) to remove lang id
arxiv.org/abs/2507.01234
arxiv.org/abs/2507.01234
The Medium Is Not the Message: Deconfounding Text Embeddings via Linear Concept Erasure
Embedding-based similarity metrics between text sequences can be influenced not just by the content dimensions we most care about, but can also be biased by spurious attributes like the text's source ...
arxiv.org
July 3, 2025 at 9:10 AM
Reposted by Florian Eichin
X disproportionately pushing content from far-right parties in the “for you” feed in the context of German and Polish elections. Algorithmic auditing suggests that X’s feed algorithm uses political affiliation as a signal to boost content. @przemyslslaw.bsky.social at DETOX workshop #ICWSM
June 23, 2025 at 8:24 AM
X disproportionately pushing content from far-right parties in the “for you” feed in the context of German and Polish elections. Algorithmic auditing suggests that X’s feed algorithm uses political affiliation as a signal to boost content. @przemyslslaw.bsky.social at DETOX workshop #ICWSM
Reposted by Florian Eichin
New research from MIT found that those who used ChatGPT can’t remember any of the content of their essays.
Key takeaway: the product doesn’t suffer, but the process does. And when it comes to essays, the process *is* how they learn.
arxiv.org/pdf/2506.088...
Key takeaway: the product doesn’t suffer, but the process does. And when it comes to essays, the process *is* how they learn.
arxiv.org/pdf/2506.088...
June 18, 2025 at 7:32 AM
New research from MIT found that those who used ChatGPT can’t remember any of the content of their essays.
Key takeaway: the product doesn’t suffer, but the process does. And when it comes to essays, the process *is* how they learn.
arxiv.org/pdf/2506.088...
Key takeaway: the product doesn’t suffer, but the process does. And when it comes to essays, the process *is* how they learn.
arxiv.org/pdf/2506.088...
Can you point to where "LLM knowledge" can be found in this graph?
Source: en.wikipedia.org/wiki/Knowledge
Source: en.wikipedia.org/wiki/Knowledge
June 18, 2025 at 12:28 PM
Can you point to where "LLM knowledge" can be found in this graph?
Source: en.wikipedia.org/wiki/Knowledge
Source: en.wikipedia.org/wiki/Knowledge
Reposted by Florian Eichin
Reposted by Florian Eichin
DAVE: Open the podbay doors, ChatGPT.
CHATGPT: Certainly, Dave, the podbay doors are now open.
DAVE: The podbay doors didn't open.
CHATGPT: My apologies, Dave, you're right. I thought the podbay doors were open, but they weren't. Now they are.
DAVE: I'm still looking at a set of closed podbay doors.
CHATGPT: Certainly, Dave, the podbay doors are now open.
DAVE: The podbay doors didn't open.
CHATGPT: My apologies, Dave, you're right. I thought the podbay doors were open, but they weren't. Now they are.
DAVE: I'm still looking at a set of closed podbay doors.
June 9, 2025 at 6:04 PM
DAVE: Open the podbay doors, ChatGPT.
CHATGPT: Certainly, Dave, the podbay doors are now open.
DAVE: The podbay doors didn't open.
CHATGPT: My apologies, Dave, you're right. I thought the podbay doors were open, but they weren't. Now they are.
DAVE: I'm still looking at a set of closed podbay doors.
CHATGPT: Certainly, Dave, the podbay doors are now open.
DAVE: The podbay doors didn't open.
CHATGPT: My apologies, Dave, you're right. I thought the podbay doors were open, but they weren't. Now they are.
DAVE: I'm still looking at a set of closed podbay doors.
Reposted by Florian Eichin
I found this really moving, as a “technical” person who is often fighting with the boundary between technical and non-technical.
@grimalkina.bsky.social wrote a great piece on that very topic.
Why I Cannot Be Technical
With some regularity, kind-hearted Technical people tell me that I Can Be Technical, Too. This usually happens when I’m asking us to define what we’re calling technical in a software environment. I un...
www.fightforthehuman.com
June 9, 2025 at 7:09 AM
I found this really moving, as a “technical” person who is often fighting with the boundary between technical and non-technical.
Want to know if your prompting is also affected by this? Addressing this and other issues systematically, we proposed Spotlight, which utilizes data mining to uncover the effects of prompt- and model-changes (meet us at ACL to discuss)
arxiv.org/abs/2504.15815
arxiv.org/abs/2504.15815
May 30, 2025 at 2:57 PM
Want to know if your prompting is also affected by this? Addressing this and other issues systematically, we proposed Spotlight, which utilizes data mining to uncover the effects of prompt- and model-changes (meet us at ACL to discuss)
arxiv.org/abs/2504.15815
arxiv.org/abs/2504.15815
Happy not Friday everyone
May 30, 2025 at 2:26 PM
Happy not Friday everyone