KInIT
@KInIT_sk
Followers
229
Following
757
Media
649
Statuses
1K
KInIT is an independent, non-profit institute dedicated to intelligent technology research. We bring together experts in different areas of computer science.
Slovakia
Joined April 2021
๐ค A key part of our lorAI project is strong collaboration among partners: @KInIT_sk, @AdaptCentre, @DFKI and Centre for Research & Technology Hellas (CERTH). ๐๏ธ During the lorAI kickoff event, we discussed why projects like this are important.
0
2
2
๐ฏ The lorAI Project aims to transform @KInIT_sk into a leading center for research and innovation in the field of low-resource AI. ๐ค Hear what @AdaptCentre, @DFKI and Centre for Research & Technology Hellas (CERTH) wish for the future of the project!
0
1
2
๐ โจ Stay tuned for the latest achievements and updates from the lorAI Project! Follow project's profile ๐ฒ #lorai
๐ก Welcome to lorAI. Launched in May on a mission to make AI more accessible, sustainable & purposeful. The journey has just begun. ๐ ๐ Discover more at https://t.co/5ETuerd7kz
#lorAI
0
0
0
The #DisAIAMPLIFIED project is making great strides ๐ โ
Validated LLMs in fact-check retrieval โ
Released first dataset โ
Probing multimodal models โ
Papers under review More to come, stay tuned! ๐
0
1
0
๐ก How can we make online recommendations not just smart, but trustworthy? With the #HERMES project & ZlavaDna, weโre running workshops ๐ค to shape the next gen of recommender systems โ balancing the needs of customers ๐ฅ, businesses ๐ผ & beyond. #hermesproject
0
0
1
What a journey it was! ๐
Last physical team meeting - and last group photo of a fantastic team! One month to go to wrap up and complete our work. Results will live on. Check out and keep an eye on https://t.co/LxLnIuZQk3, also after the project formally ends on 31 Oct 2025.
1
1
7
๐ก @KInIT_sk researchers created MultiSocial, a new dataset to test how well tools can spot #AI-generated text on social media. ๐ Fine-tuned detectors adapt best, and platform & language choice affects performance. ๐ https://t.co/Qx3d8gUiq6
#VIGILANTProject #DigitalSecurity
0
2
4
๐ผ๏ธ NEW BLOG: Final part of our #DisAIAMPLIFIED series! How can images boost fact-checking? ๐ท Visuals = key evidence ๐ค Text + multimodal + LLMs ๐ OCR when text hides in images ๐ https://t.co/rISThL6Cjl
#FactChecking #Misinformation #TrustworthyAI #NLP
0
0
0
๐ Our Research Engineer Philipp Miotti joined Peter Straลพovec (PayOut) at #AIConference2025 by Deutsche Telekom IT Solutions Slovakia to present Fraudnet - an AI + federated learning solution for real-time fraud prevention. #KInIT #FraudNet #FederatedLearning #FraudPrevention
0
0
0
โณ Almost time for #RecSys2025! โจ Our CEO Mรกria Bielikovรก & lead researcher Michal Kompan are co-chairs, and PhD student Santiago de Leon Martinez will run a workshop. ๐ Big thanks to the organizers for a great program. See you in Prague! ๐ ๐ https://t.co/SOeiA6DWFS
0
0
0
Join us in two weeks in Prague at CEDMO Cafรฉ. Our Lead and Researcher Jakub ล imko is hosting "Fact-checking and AI Tools under Pressure" panel. ๐๏ธ See you on 23rd September 2025. ๐ Register via link in comments ๐
๐ข Introducing our "Espresso Panel: Concentrated Narratives - Fact-checking and AI Tools Under Pressure"! Like a strong espresso, this panel delivers intense discussion on the AI-powered tools developed to detect disinformation or deepfakes and counter them. ๐๏ธ Join us on 23
0
1
2
๐ Great resource from @EBU_HQ : a handbook on C2PA (Coalition for Content Provenance and Authentication) and its role in building trust and countering disinformation. ๐ Also, don't forget to check possible interlinking with @veraai_eu tools ๐ง
Published today, written by @EBU_HQ colleagues: "Handbook on the legal and ethical obligations to developers and deployers of AI-based fact-checking tools". Available for reading and download right here. ๐ https://t.co/UB64bsKtly
0
1
1
๐ค๐ Spotting #AI text on social media? @KInIT_sk's MultiSocial dataset (470k posts, 22 languages, 5 platforms) tests how well tools detect AI vs human content. Results show fine-tuned models work bestโeven small ones! ๐ก ๐Check it out: https://t.co/Ogg3a08O5g
#Disinformation
0
4
5
Fact-checking isnโt just about text ๐ Visual data can make a big difference โ even across languages. ๐ Our #DisAI paper โMultimodal and Multilingual Fact-Checked Article Retrievalโ is out now. Read here ๐ https://t.co/WFidZ0eaAa
#ResearchPaper #MultimodalAI #FactChecking
0
1
1
Are you aware of the challenges faced by low-resource languages against disinformation? #EUfunded research is changing that narrative. Tune in to #CORDIScovery podcast to learn how the EU is empowering fact-checkers & tackling disinformation! ๐ https://t.co/bx6O9BFL3x
0
4
8
4๏ธโฃ Thatโs why AI ethics is understood as the ethics of developing and deploying AI systems. Itโs about the values guiding the humans who create them, and how those values can be translated into concrete technological solutions.
0
0
0
3๏ธโฃ ๐ and most crucially many ethical challenges arise already during development, not just when AI is deployed.
0
0
0
2๏ธโฃ Why do we speak about โAI ethicsโ as a separate field? Because AI systems: ๐ impact large numbers of people at once, ๐ often operate in ways we canโt fully see or explain (the โblack boxโ problem),
0
0
0
1๏ธโฃ While this is an important aspect, it is equally crucial to focus on the people who design and implement these systems and to support them in making the right decisions.
0
0
0