PAN at CLEF 2025

Shared Tasks

Important Dates

  • January 31, 2025: Training Data Released
  • May 23, 2025: Software submission deadline
  • May 30, 2025: Participant paper submission
  • June 27, 2025: Peer review notification
  • July 07, 2025: Camera-ready participant papers submission
  • September 09-12, 2025: Conference

The time of all deadlines is Midnight CEST.

Keynotes

Jussi Karlgren
Cultural factors in language use and generative AI
Jussi Karlgren
University of Helsinki

A generative language model can be used to generate task- and situation-appropriate material to fit some purpose. Training a model for this purpose is done through post-training such as fine-tuning, alignment, instruction, and human feedback.

Doing this—imposing constraints on what a model can generate—is not a value-neutral process! These constraints, while they can be imposed with reference to universally accepted and agreed-upon human values such as courtesy, helpfulness, and empathy, are instantiated variously across languages and cultures. Thinking about this is something that intersects interestingly with interests of PAN!

Jussi Karlgren is a Principal AI Scientist at Silo AI, which is now a part of AMD, and a docent of language technology at Helsinki University. He has co-founded the text analysis company Gavagi and has held various research-related roles at major companies and organizations such as Spotify, IBM Nordic Laboratories, Columbia University, Xerox PARC, New York University, Yahoo! Research in Barcelona, and Stanford University. Jussi is known for his work in computational linguistics and stylometry and for first formulating the notion of a recommender system.

Read more… Read less…

Program

The program is still preliminary and subject to (minor) change!

PAN's program is part of the CLEF 2025 conference program. All times are Central European Summer Time - CEST.

The PAN lab will take place in the Salón de Actos of the Facultad de Educación building.
Wednesday, September 10
11:30-13:15 CLEF Session: Lab Overviews (PAN, EXIST, SimpleText, QuantumCLEF) – Salón de Actos
14:15-15:45 Keynote and Lab Session, Chair: Martin Potthast – Salón de Actos
14:15-14:45 Keynote: Cultural factors in language use and generative AI
Jussi Karlgren
14:45-15:15 Overview: Multi-Author Writing Style Analysis at PAN 2025
Eva Zangerle, Maximilian Mayerl, Martin Potthast and Benno Stein
15:15-15:30 Style Change Detection Using Graph and Structural Linguistic Features for Multi-Author Writing Analysis
Ioana-Roxana Boriceanu and Andra-Elena Băltoiu
15:30-15:45 Team "better_call_claude": Style Change Detection using a Sequential Sentence Pair Classifier
Gleb Schmidt, Johannes Römisch, Mariia Halchynska, Svetlana Gorovaia and Ivan Yamshchikovgit
15:45-16:30 Coffee Break and Poster Session – Canteen and Ground Floor
16:30-18:00 Lab Session, Chair: Janek Bevendorff – Salón de Actos
16:30-16:45 Overview: Generative Plagiarism Detection
André Greiner-Petter, Maik Fröbe, Jan Philip Wahle, Terry Ruas, Bela Gipp, Akiko Aizawa and Martin Potthast
16:45-17:15 Overview: Multilingual Text Detoxification at PAN 2025
Daryna Dementieva, Vitaly Protasov, Nikolay Babakov, Naquee Rizwan, Ilseyar Alimova, Caroline Brune, Vasily Konovalov, Arianna Muti, Chaya Liebeskind, Marina Litvak, Debora Nozza, Shehryaar Shah Khan, Sotaro Takeshita, Natalia Vanetik, Abinew Ali Ayele, Frolian Schneider, Xintog Wang, Seid Muhie Yimam, Ashraf Elnagar, Animesh Mukherjee and Alexander Panchenko
17:15-17:30 nikita.sushko at TextDetox CLEF 2025: Exploring A Sage-T5-like Approach For Text Detoxification
Alexandr Voronin, Daniil Moskovsky and Nikita Sushko
17:30-17:45 SVATS at PAN 2025 TextDetox: Can Small Models Outperform Large Ones in Text Detoxification?
Vladislav Kozlovskiy, Alexander Ploskin, Sameer Tantry, Tatyana Matveeva and Sofya Savelyeva
Thursday, September 11
14:15-15:45 Lab Session, Chair: Benno Stein – Salón de Actos
14:15-14:35 Overview: “Voight-Kampff” Generative AI Authorship Verification at PAN and ELOQUENT 2025 (Subtask 1)
Janek Bevendorff, Jussi Karlgren, Matti Wiegmann, Maik Fröbe, Efstathios Stamatatos, Martin Potthast and Benno Stein
14:35-14:45 DivEye at PAN 2025: Diversity Boosts AI-Generated Text Detection
Advik Basani and Pin-Yu Chen
14:45-14:55 Team Nexus Interrogators at PAN: Voight-Kampff Generative AI Detection
Samiya Ali Zaidi, Huzaifah Tariq Ahmed, Sarrah Ali Akbar, Ziaullah Shakeel, Faisal Alvi and Abdul Samad
14:55-15:10 mdok of KInIT: Robustly Fine-tuned LLM for Binary and Multiclass AI-Generated Text Detection
Dominik Macko
15:10-15:25 Human or Not? Light-Weight and Interpretable Detection of AI-Generated Text
Maximilian Seeliger, Patrick Styll, Moritz Staudinger and Allan Hanbury
15:25-15:45 Overview: Human-AI Collaborative Text Detection at PAN 2025 (“Voight-Kampff” Subtask 2)
Yuxia Wang, Akim Tsivgun, Jinyan Su, Zhuohan Xie, Mervat Abassy, Jonibek Mansurov, Rui Xing, Minh Ngoc Ta, Kareem Ashraf Elozeiri, Tianle Gu, Raj Vardhan Tomar, Jiahui Geng, Ekaterina Artemova, Artem Shelmanov, Nizar Habash, Iryna Gurevych
15:45-16:30 Coffee Break – Canteen and Ground Floor
16:30-18:00 Lab Session, Chair: Janek Bevendorff – Salón de Actos
16:30-16:40 Unibuc - NLP at "Voight-Kampff" Generative AI Detection PAN 2025
Teodor Marchitan, Claudiu Creanga and Liviu P. Dinu
16:40-16:50 DBG: Human-AI Collaborative Text Classification with DeBERTa-enhanced Contextual and Geometric Attention
Tufeng Xian, Yong Zhong, Fen Liu, Meifang Xie, Qiyuan Sun, Miaoji Zheng, Weidong Wu and Zhiliang Zhang
16:50-17:00 Overview: Voight-Kampff at ELOQUENT 2025
Jussi Karlgren
17:00-17:10 The Distinctive Data-Centric Approach for the Voight Kampff Task
Rohit R. Gunti
17:10-17:20 Literal Re-translation as a Method for AI Text Disguise and Detection Evasion
Poojan Vachharajani
17:20-17:50 Overview: The Robustness and Consistency Task of the ELOQUENT 2025 Lab for Evaluating Generative Language Model Quality
Jussi Karlgren, Marie Isabel Engels, Maria Barrett, Rohit Raj Gunti, Mohanna Hoveyda, Bruno Nadalic Sotic, Jaap Kamps, Mika Koistinen, Elaine Zosa
17:50-18:00 Closing

Participation Modalities

To participate at PAN, first register for your task of choice at CLEF, then follow the instructions below.

Submission

We use TIRA for all submissions to PAN. Please go to tira.io, create an account, and register for the individual tasks you want to participate in. You need to submit your software or your results via TIRA. You can find all submission guides in TIRA's forum.

Data

You can download PAN's datasets here. For details, please check the individual task's website.

Evaluation and Baseline Code

All code used at PAN is published at GitHub. You can find all validators, evaluators, and baselines there.

Software Submissions

PAN promotes reproducible science with software submissions. Please prepare and submit your software as Docker image(s). You can find guides and examples in the resources linked above. Some tasks allow only software submissions and only release the test data after the conference.

Paper Submission and Presentation

PAN is co-located with CLEF 2025 in Madrid, Spain. Every participant is expected to write a notebook paper describing their approach to CLEF (published at CEUR-WS, which is indexed by DBLP). At the CLEF conference, all submissions will be presented as talks or posters. CLEF will be a hybrid conference.