Our work (co-authors: Paweł Skórzewski, Marcin Sowański, Tomasz Zietkiewicz) on using back transcription for evaluating robustness of #NLU models to speech recognition errors was featured on Samsung Research Blog research.samsung.com/blog/Back-Tran… #AI #NLProc Department of Artificial Intelligence AMU Poznan Samsung Research

Piotr Jabloński, our PhD student, is attending the winter school during the Northern Lights Deep Learning Conference in Tromsø, where workshops are being conducted by researchers from the University of Cambridge and the University of Tromsø. Piotr's participation is made possible thanks to IDUB UAM funds

Members of our Artificial Intelligence department are participating in aUPaEU workshop on the Agora concept.

In exactly 20 minutes Marek Kubis Paweł Skórzewski Tomasz Zietkiewicz and Marcin Sowański will speak about Back Transcription as a Method for Evaluating Robustness of NLU Models to Speech Recognition Errors. Join us online or in person. We start at 11.00 am CET wmi.amu.edu.pl/zycie-naukowe/…

We're starting our seminar where Preslav Nakov from Mohamed bin Zayed University of Artificial Intelligence will speak about "Factuality Challenges in the Era of Large Language Models". Join us right now at wmi.amu.edu.pl/zycie-naukowe/…

Jakub Barylski and Andrzej Zieleziński are showcasing their project on "Use of language models to predict viral and bacterial gene function". Catch their presentation at 11am (CET), in-person or online wmi.amu.edu.pl/zycie-naukowe/…

Benchmark Intended Grouping of Open Speech (BIGOS) is a system developed by Michał Juńczyk. In less than 5 minutes, Michał will present his progress on this project. Join us on-site or online wmi.amu.edu.pl/zycie-naukowe/…

We're starting our seminar where prof. UKW dr hab. Edward Jacek Gorzelańczyk from Uniwersytet Kazimierza Wielkiego w Bydgoszczy will speak about Smart cyber-psychiatry office. Join us on-site or online wmi.amu.edu.pl/zycie-naukowe/…

Today, Mateusz Tylka will present our evaluation system AMUEval (previously Gonito), rewritten in Python and distributed under an Open Source license. Join us at B1-7/8 or online at wmi.amu.edu.pl/zycie-naukowe/…

POLygraph - our team Paweł Skórzewski Marek Kubis Filip Graliński Piotr Jabłoński and others, prepared a unique resource for fake news detection in Polish which will be presented at WASSA 2026 workshop during ACL 2025 2024 🚀🚀 Paper available at arxiv.org/abs/2407.01393 📄

Today, prof. Jakub D. Rybka (Center for Advanced Technology AMU) spoke about "Potential applications of AI in artificial meniscus design". Don't miss out our future seminars wmi.amu.edu.pl/zycie-naukowe/…

"Scintilla: Simulating Combustible Vegetation for Wildfires" is a paper presented at SIGGRAPH 2024 by Andrzej Kokosza, Miłosz Makowski, Wojtek Pałubicki, and others, which introduces a new method for simulating wildfires. Check out the full paper at dl.acm.org/doi/10.1145/36…

Today was tle last day of workshops at ACL 2025 conference. Yesterday at WASSA 2026, Paweł Skórzewski and Piotr Jabłoński represented our department with paper „POLygraph: Polish Fake News Dataset” check full paper at aclanthology.org/2024.wassa-1.2…

Some lessons (I) learnt preparing the data mixture for Snowflake Arctic LLM 👨‍🍳

LLM Bielik v2 on our internal benchmark, based on Polish educational and professional tests, achieves an accuracy score of 58.03%. This is a noticeable improvement over the 41,51% in v.0.1. Congratulations to the entire SpeakLeash team. More extensive results coming 🔜

A joint study by Department of Artificial Intelligence AMU Poznan researchers and Samsung Electronics Polska engineers was presented at FedCSIS 2024. The paper investigates the impact of augmenting spoken language corpora with domain-specific synthetic samples. arxiv.org/abs/2406.07090

In exactly 30 minutes, we're starting our weekly seminar titled "HANOI corpus. A unique way of recording information, or notation in consecutive translation" with PhD Anna Jelec from Faculty of English AMU (Polish Language) Join us on site or remotely wmi.amu.edu.pl/zycie-naukowe/…

We're starting our weekly seminar titled "AI to help domain bibliographies (and vice versa)" with PhD Cezary Rosiński and MSc. Dariusz Perliński from Polish Academy of Sciences(Polish Language) Join us on site or remotely wmi.amu.edu.pl/zycie-naukowe/…

GPT-5-Pro solved, in just 15 minutes (without any internet search), the presentation problem known as “Yu Tsumura’s 554th Problem.” arxiv.org/pdf/2508.03685 This is the first model to solve this task completely. I expect more such results soon — the model demonstrates a strong