martin-potthast.com
Professor at the University of Kassel, https://hessian.AI, and https://ScaDS.AI. Member of @webis.de
Research in information retrieval #IR, natural language processing #NLP, and artificial intelligence.
26 posts
2,432 followers
1,288 following
Regular Contributor
Active Commenter
comment in response to
post
We share your concern that LLMs could be prompted to generate responses that are biased in favor of certain products. That is why we are currently organizing a shared task on detecting advertisements in the responses of RAG-based search engines: bsky.app/profile/webi...
comment in response to
post
🙋♂️
comment in response to
post
Important Dates
----------------------
now Training Data Released
May 23, 2025 Software submission
May 30, 2025 Participant paper submission
June 27, 2025 Peer review notification
July 07, 2025 Camera-ready participant papers submission
Sep 09-12, 2025 Conference
comment in response to
post
4. Generative Plagiarism Detection.
Given a pair of documents, your task is to identify all contiguous maximal-length passages of reused text between them.
pan.webis.de/clef25/pan25...
comment in response to
post
3. Multi-Author Writing Style Analysis.
Given a document, determine at which positions the author changes.
pan.webis.de/clef25/pan25...
comment in response to
post
2. Multilingual Text Detoxification.
Given a toxic piece of text, re-write it in a non-toxic way while saving the main content as much as possible.
pan.webis.de/clef25/pan25...
comment in response to
post
1. Voight-Kampff Generative AI Detection.
Subtask 1: Given a (potentially obfuscated) text, decide whether it was written by a human or an AI.
Subtask 2: Given a document collaboratively authored by human and AI, classify the extent to which the model assisted.
pan.webis.de/clef25/pan25...
comment in response to
post
True. No one will bother to ask the same question more than once, but will take the answer that comes as a given and as a convenient starting point to move on.
This has been the predictable result since Google introduced featured snippets: webis.de/publications...
comment in response to
post
On second thought, it does at least hit the the right ballpark.
Purple is hardly discernible from black in the pictures, and the perspective does make red look a little bit like it has another piece secured.
comment in response to
post
Apparently, it can be made to second-guess itself quite easily.
4/4
comment in response to
post
Yellow won. GPT does not see it.
Potentially, the perspective plays a role.
3/4
comment in response to
post
... who move to your own position(s).
Despite being almost entirely a game of luck, it's surprisingly funny, known to be just as upsetting on occasion, and reasonably well balanced until the end.
Here, yellow is in the lead. but GPT recognizes it only partially. 2/4
comment in response to
post
This may help: netspeak.org
webis.de/publications...
comment in response to
post
More IR starter packs!
go.bsky.app/MXPJoTn
comment in response to
post
🙋♂️
comment in response to
post
🙋♂️
comment in response to
post
D!
comment in response to
post
Really exciting work. I'm happy to see more work in the related to our trigger warning research popping up. Perhaps you'll find that interesting, too: webis.de/publications...
comment in response to
post
Can you add me to the list as well?
comment in response to
post
This might be interesting in relation to this paper: bsky.app/profile/sung...
comment in response to
post
Android or iOS, and how did you do that?
comment in response to
post
You can add me, please.