Ion Androutsopoulos retweeted
🔌 #AthNLP2026 welcomes @omilialtd back as Silver Sponsor, continuing its support from last year.
Omilia develops conversational AI systems that enable more natural interaction between humans and machines.
This collaboration supports the #NLP & #AI community at AthNLP.
#Omilia
Ion Androutsopoulos retweeted
🚀 AthNLP 2026 welcomes @EndeavorGr back as Platinum Sponsor, continuing its support from last year.
Through its commitment to innovation ecosystems, Endeavor supports AthNLP’s mission to bring together the NLP & AI community.
#AthNLP2026 #NLP #AI #MachineLearning #Endeavor
Ion Androutsopoulos retweeted
#AthNLP2026 kicked off this morning at NCSR "Demokritos" 🇬🇷
Over the next week, students & researchers from across Europe and beyond gather for lectures, workshops, and discussions with leading experts in #NLP and #AI
Welcome to all our participants, excited to have you here! 👋
Ion Androutsopoulos retweeted
The MSc in Artificial Intelligence & Data Science at the @athens_and continues its support for #AthNLP2026, following its involvement in last year’s edition.
This collaboration reflects a shared commitment to education and research in #NLP, #MachineLearning, and #AI.
Ion Androutsopoulos retweeted
🎤 #AthNLP2026 Speaker Announcement
@ikekong (@HKUniversity )
Focus: neural architectures, long-context modeling & reasoning in LLMs
🎯 Talk: Text Diffusion Models: What’s Not an Autoregressive LLM?
📌 athnlp.github.io/2026/
#NLP #AI #MachineLearning #LLM #DeepLearning
Ion Androutsopoulos retweeted
🎤 #AthNLP2026 Speaker Announcement
@IAugenstein (@uni_copenhagen)
Focus: fair & accountable #NLP, with emphasis on explainability, factuality, and bias detection.
🎯 Talk: Understanding LLMs’ Knowledge Utilisation
📌 athnlp.github.io/2026/
#AI #LLM #ExplainableAI
Ion Androutsopoulos retweeted
✈️ Proud to welcome @aegeanairlines
back as Platinum Sponsor & Official Air Carrier of #AthNLP2026.
For a third consecutive year, AEGEAN supports a more international, accessible, and connected NLP & AI community.
#NLP #AI #MachineLearning #LLM
Ion Androutsopoulos retweeted
🎤 AthNLP 2026 Speaker Announcement
#AthNLP2026 welcomes @annargrs (@koebenhavns_uni ).
Focus: interpretability, robustness & societal impact of LLM-based NLP systems.
🗣️ Talk: LLMs and Factuality
🔗 athnlp.github.io/2026/speake…
#NLP #AI #LLM #MachineLearning
Ion Androutsopoulos retweeted
🎤 #AthNLP2026 Speaker Announcement
@Ryan McDonald AI consultant & investor)
Expert in ML for NLP, working on autonomous agents & scientific simulation.
🎯 Talk: Classification
#AthNLP #NLP #AI #MachineLearning #Classification
Ion Androutsopoulos retweeted
🎤 @delliott joins #AthNLP2026!
Associate Professor and Villum Young Investigator at the @uni_copenhagen, working on #multimodal #AI, visual representations of text, and multilingual language processing.
🗣️ Talk: “Vision and Language”
For more: athnlp.github.io/2026/speake… #NLP
Ion Androutsopoulos retweeted
🎤 AthNLP 2026 Speaker Announcement
#AthNLP2026 welcomes Dr Dong Nguyen @UniUtrecht
Focus: #fairness, robustness & language variation in #NLP, with data-centered, real-world applications.
🗣️ Talk: RNN + Transformers
🔗 dongnguyen.nl/
📌 Apply: athnlp.github.io/2026/cfp.ht…
Ion Androutsopoulos retweeted
#AthNLP2026 Speaker Announcement
AthNLP 2026 welcomes Julia Kreutzer (@Cohere_Labs)
Focus: multilingual #LLMs and machine translation, with a focus on underrepresented languages.
Talk: RLHF From Scratch
🌐 athnlp.github.io/2026/speake…
📌 athnlp.github.io/2026/cfp.ht…
#NLP #AI
Ion Androutsopoulos retweeted
🚀 This September, Athens becomes a hub for cutting-edge NLP & AI research 🇬🇷
AthNLP (Sept 2–8) brings students & researchers together for a week on NLP, ML & AI.
LLMs • Multimodal AI • Modern NLP systems
📌 athnlp.github.io/2026/cfp.ht…
#AthNLP #NLP #AI #MachineLearning #LLM
Ion Androutsopoulos retweeted
📅 #AthNLP2026 – Preliminary Schedule is now available!
A full week of lectures and labs on #NLP, #AI, #LLMs, and Multimodal systems, covering topics such as transformers, sequence modeling, machine translation, LLM agents, and evaluation.
📌athnlp.github.io/2026/schedu…
Η Ακαδημία Αθηνών διοργανώνει εκδήλωση με θέμα την Τεχνητή Νοημοσύνη, την Παρασκευή 15 Μαΐου 2026, 17.00-21:10. Η συμμετοχή είναι ελεύθερη. Δυνατότητα διαδικτυακής παρακολούθησης. Πρόγραμμα και περισσότερες πληροφορίες:
academyofathens.gr/el/news/s…
Ion Androutsopoulos retweeted
📢 #AthNLP26 CFP is open!
📅 Sept 2–8, Athens 🇬🇷
🗓️ Deadline: May 31
#NLP • #AI • #LLMs • #ML
📌 Apply: ijerm0co.forms.app/athens-nl…
📎 athnlp.github.io/2026/
Stay tuned for more!
Most of the videos and slides of the AthensNLP 2025 summer school (September 2025) are now online (athnlp.github.io/2025/presen…)! Stay tuned for AthensNLP 2026 (athnlp.github.io/2026/)!
Ion Androutsopoulos retweeted
Excited to announce our first project from the HuggingLegal community: the Gemini-3-Benchmarkathon!
Gemini-3 achieved the top spot on most major benchmarks last week, but how well does it know the law? Unfortunately, most model providers don't evaluate on law-specific benchmarks. So while we have a good idea of how good new models are at coding, we are pretty much in the dark about their lawyering abilities.
This is why we ran a vibe-check on six diverse datasets from Greek bar exams over Indian law questions to Swiss university law exams among others. So what do the vibes say?
AA-Omniscience: 6/10 – high competence but unreliable
LegalBench: 9.5/10 – almost perfect answers
GreekBarBench: 9/10 – impressive long-context reasoning without hallucinations
IndianLawQA: 8.5/10 – The task feels precise, clean, and insightful, revealing how reliably Gemini-3 handles high-precision statutory queries (including new BNSS 2023 codes) while most models typically hallucinate in this domain.
WilfulMisconduct: 8/10 – strong logic, missed binding precedent
LEXam: 7/10 – competent but overly confident
Gemini-3 is extremely strong at many legal tasks and often performs at or above the level of very good human lawyers. However, it still makes serious mistakes: it often answers confidently even when it does not know the right legal rule or fact, instead of saying “I don’t know.” Because of this overconfidence and some failures on complex reasoning and precedents, it cannot safely replace human lawyers and still needs expert oversight.
This is a great community effort, thanks for the collaboration Robert Scholz, @5vsas, Ernest Beta, @odychlapanis, @adhipba, Matteo Bürgler, Sophie Franco, Chu Fei Luo, @samdahan06!
Find the link to the article below.
Ion Androutsopoulos retweeted
If you're interested in low-resource or domain-specific evaluation checkout this cool dataset by @odychlapanis and Dimitrios Galanis and @nikaletras and @ionandrou!
GreekBarBench is a benchmark designed to evaluate LLMs on challenging legal reasoning questions across five different legal areas from the Greek Bar exams, requiring citations to statutory articles and case facts.








