iiinova — where your digital self keeps evolving

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 09/07/2026

Nalin Kumar & @tuetschek.bsky.social presented Modular Monolingual Adaptation using Pretrained Language Models aclanthology.org/2026.acl-ind... with their tricks for saving parameters and improving performance when fine-tuning pretrained LMs for low-resource languages.

032

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 08/07/2026

Reasoning Gets Harder for LLMs Inside A Dialogue aclanthology.org/2026.acl-lon... by @ivankartac.bsky.social, M. Lango & @tuetschek.bsky.social Same reasoning tasks, isolated vs. inside a dialogue: 9 LLMs 📉 consistently do worse once conversation, roles & tool-use requirements enter the picture.

022

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 07/07/2026

One-step Nonautoregressive Natural Language Generation with Shortcut Flow Matching Models by Jędrzej Warczyński, Mateusz Lango & @tuetschek.bsky.social aclanthology.org/2026.acl-sho... One-step generation that actually works: BLEU more than doubles vs classic flow matching.

032

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 07/07/2026

Thesis Proposal by Kristýna Onderková: Intentional Inference for Insight Generation aclanthology.org/2026.acl-srw... Why do LLM insights feel shallow? Often, it is unstated assumptions that fill gaps in the task. This thesis makes models surface them & push toward deeper, more trustworthy reasoning.

021

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 05/07/2026

UFAL-CUNI at SemEval-2026 Task 11: An Efficient Modular Neuro-symbolic Method for Syllogistic Reasoning by @ivankartac.bsky.social, Honza Bronec, Kristýna Onderková, @zdenekkasner.cz, Mateusz Lango & @tuetschek.bsky.social aclanthology.org/2026.semeval... 💪 4B model + prover beats zero-shot LLMs

051

Reposted by Ondrej Dusek

INLG 2026 @inlg.bsky.social · 25/06/2026

The 2nd Call for Papers for #INLG2026 is out! Two ways to submit: 📄 Direct submissions, archival and not, by 15 July: openreview.net/group?id=acl... 💍 ARR commitments by 5 August: openreview.net/group?id=acl... ——— 📍Utrecht, NL — Oct 17–21, 2026, right before EMNLP 2026.inlgmeeting.org/calls.html

openreview.net

INLG 2026 Conference Direct Submissions

Welcome to the OpenReview homepage for INLG 2026 Conference Direct Submissions

087

Reposted by Ondrej Dusek

Darth Putin @darthputinkgb.bsky.social · 16/05/2026

Never in the field of human endeavour has so much been fucked up for so many by so few.

291351311

Reposted by Ondrej Dusek

Ivan Habernal @ivanhabernal.bsky.social · 16/05/2026

Looks totally legit. Right? RIGHT? (these two fraudster accounts compromised our PrivateNLP ACL workshop - wasted time of reviewers, chairs, never replied back... We were not the only ones, so I asked OpenReview to ban their accounts for scientific fraud)

153

Reposted by Ondrej Dusek

Ivan Kartáč @ivankartac.bsky.social · 13/04/2026

Happy to see this accepted at #ACL2026 main!

042

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 29/03/2026

AnimatedLLM: Explaining LLMs with Interactive Visualizations by @zdenekkasner.cz and @tuetschek.bsky.social aclanthology.org/2026.teachin... AnimatedLLM lets you explore how LLMs work step by step. Right in your browser, no setup needed. Great for teaching or self-study! 🧵🤖 👉 animatedllm.github.io

141

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 28/03/2026

LLMs as Span Annotators: A Comparative Study of LLMs and Humans by @zdenekkasner.cz, @zouharvi.bsky.social, @patuchen.bsky.social, @ivankartac.bsky.social, K. Onderková, @oplatek.bsky.social, Dimitra Gkatzia, @saad.me.uk , @tuetschek.bsky.social & Simone Ballocu aclanthology.org/2026.mme-mai...

1134

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 28/03/2026

Keynote by @tuetschek.bsky.social on LLM evaluation: standard metrics fall short of catching subtle errors, and human evaluation lacks consistency. Span-level error annotation with LLM-as-judge ensembles ➡️ LLMs matching trained human annotators. At the LowResMT workshop.

291

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 24/03/2026

#EACL2026 in Rabat 🇲🇦 starts tomorrow and @ufal.mff.cuni.cz folks will present their research. Don't miss our presentations 👇

1132

Reposted by Ondrej Dusek

Ivan Kartáč @ivankartac.bsky.social · 24/03/2026

New preprint: arxiv.org/abs/2603.20133 Does performance on reasoning benchmarks transfer to real-world settings such as task-oriented dialogue? Not necessarily: our new benchmark tests LLMs on problems framed in both standalone and dialogue settings, and shows that dialogue makes reasoning harder.

1176

Reposted by Ondrej Dusek

Juan Diego Rodriguez @juand-r.bsky.social · 17/02/2026

Claude: For each example I can do a web search and then make a LLM call with the results... Me: Why an LLM call? Can't you just figure it out yourself? Claude: You're right, I am the LLM!

2437

Reposted by Ondrej Dusek

Kyle Lo @ ICML2026 🇰🇷 @kylelo.bsky.social · 27/01/2026

The 5th Generation, Evaluation, and Metrics (GEM) Workshop will be at #ACL2026! Call for papers is out. Topics include: 🐟 LMs as evaluators 🐠 Living benchmarks 🍣 Eval with humans and more New for 2026: Opinion & Statement Papers! Full CFP: gem-workshop.com/call-for-pap...

0227

Reposted by Ondrej Dusek

Zdeněk Kasner @zdenekkasner.cz · 29/01/2026

If you think labeling text spans with LLMs is easy, you probably have not tried it yourself (we have! 🙃). Any method you can think of – be it tagging, matching, or indexing – has flaws. In our new preprint, we tested them all 💪We also proposed how to improve one of them. arxiv.org/abs/2601.16946

2396

Reposted by Ondrej Dusek

Zdeněk Kasner @zdenekkasner.cz · 18/12/2025

Do you often find yourself explaining how LLMs work to your students, parents, kids or other teachers? AnimatedLLM can make your life easier! animatedllm.github.io #NLP #NLProc @ufal.mff.cuni.cz @tuetschek.bsky.social

animatedllm.github.io

AnimatedLLM - Explaining LLMs with Interactive Visualizations

Understand how large language models work under the hood.

182

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 11/11/2025

🔤 Pretraining Language Models with LoRA and Artificial Languages Nalin Kumar, Mateusz Lango, @tuetschek.bsky.social t aclanthology.org/2025.babylm-... Constructed artificial languages with LoRA affects language model development.

aclanthology.org

Pretraining Language Models with LoRA and Artificial Languages

Nalin Kumar, Mateusz Lango, Ondrej Dusek. Proceedings of the First BabyLM Workshop. 2025.

131

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 11/11/2025

🎓 You are an LLM teaching a smaller model everything you know: Multi-task pretraining of language models with LLM-designed study plans Wiktor Kamzela, Mateusz Lango, @tuetschek.bsky.social aclanthology.org/2025.babylm-...

aclanthology.org

You are an LLM teaching a smaller model everything you know: Multi-task pretraining of language models with LLM-designed study plans

Wiktor Kamzela, Mateusz Lango, Ondrej Dusek. Proceedings of the First BabyLM Workshop. 2025.

121

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 07/11/2025

📚 SRS-Stories: Vocabulary-constrained multilingual story generation for language learning Wiktor Kamzela, Mateusz Lango & @toonietuesday.bsky.social aclanthology.org/2025.emnlp-i... LLM stories teach vocab while reviewing learned words via Spaced Repetition-more grammatical than standard generation

131

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 07/11/2025

🤖 LLM Agents Implement an NLG System from Scratch Mateusz Lango, Ondrej Dusek aclanthology.org/2025.emnlp-i... LLM agents can autonomously build interpretable, rule-based RDF-to-text generators from scratch, combining the LLMs with the transparency and reliability of traditional rule-based systems.

aclanthology.org

LLM Agents Implement an NLG System from Scratch: Building Interpretable Rule-Based RDF-to-Text Generators

Mateusz Lango, Ondrej Dusek. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track. 2025.

131

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 07/11/2025

👥 Can Large Language Models Personalize Dialogues to Generational Styles? P. Balestrucci, @tuetschek.bsky.social, L. Anselma, A. Mazzei aclanthology.org/2025.finding... Can LLMs adapt dialogues to generational styles? We show with P-MultiWoZ that models capture patterns from Boomers to Gen Z.

aclanthology.org

Can Large Language Models Personalize Dialogues to Generational Styles?

Pier Felice Balestrucci, Ondrej Dusek, Luca Anselma, Alessandro Mazzei. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.

131

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 07/11/2025

📊 Real-World Summarization: When Evaluation Reaches Its Limits @patuchen.bsky.social , @tuetschek.bsky.social , @saad.me.uk aclanthology.org/2025.finding... For hotel highlights, metrics like word overlap surprisingly match human judgments better than complex methods. LLMs unreliable as evaluators.

aclanthology.org

Real-World Summarization: When Evaluation Reaches Its Limits

Patrícia Schmidtová, Ondrej Dusek, Saad Mahamood. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.

142

Reposted by Ondrej Dusek

SIGGEN & INLG @siggen.bsky.social · 16/09/2025

The registration page for #INLG2025 is now live! Join us in Vietnam at the Oct 29 - Nov 2 for the best conference on #NaturalLanguageGeneration 2025.inlgmeeting.org/registration... Curious to see what will be presented? Check out this list of accepted papers! 2025.inlgmeeting.org/accepted-pap...

Picture of the One Pillar Pagoda in Hanoi, a pagoda raised up over a green pond surrounded by greenery

044

Reposted by Ondrej Dusek

Svitlana Vakulenko @vendi12.bsky.social · 09/09/2025

Check out the slides from our SCAI'2025 #convsearch workshop collocated with @ijcai.org #IJCAI2025 on LLMs, retrieval & QA, recommendations, negotiations, evaluation and transparency scai.info/scai-2025 @patuchen.bsky.social @maik-froebe.bsky.social @tuetschek.bsky.social @mila-quebec.bsky.social

scai.info

SCAI 2025

Online Event on Search-Oriented Conversational AI.

093

Reposted by Ondrej Dusek

Ivan Kartáč @ivankartac.bsky.social · 23/08/2025

Our paper "OpeNLGauge: An Explainable Metric for NLG Evaluation with Open-Weights LLMs" has been accepted to #INLG2025 conference! You can read the preprint here: arxiv.org/abs/2503.11858

142

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 31/07/2025

FreshTab: Sourcing Fresh Resources for Table-to-Text Generation Evaluation by @navitas.bsky.social, ‪@oplatek.bsky.social‬, ‪@zdenekkasner.bsky.social‬, @tuetschek.bsky.social .bsky.social‬

161

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 31/07/2025

ReproHum #0669-08: Reproducing Sentiment Transfer Evaluation by @navitas.bsky.social, M. Lango, @patuchen.bsky.social, @tuetschek.bsky.social Challenge to reproduce human evaluations from NLP papers, testing the reproducibility of evaluation studies

161

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 31/07/2025

OpeNLGauge: An Explainable Metric for NLG Evaluation with Open-Weights LLMs by @ivankartac.bsky.social, M. Lango, @tuetschek.bsky.social arxiv.org/abs/2503.11858 Open-source NLG evaluation metric that explains errors and matches human judgments without proprietary models

171

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 28/07/2025

#ACL2025NLP in Vienna 🇦🇹 starts today with 23 🤯 @ufal-cuni.bsky.social folks presenting their work both at the main conference and workshops. Check out our main conference papers today and on Wednesday 👇

1228

Reposted by Ondrej Dusek

Mark Riedl @markriedl.bsky.social · 24/07/2025

ICML found hidden prompts in accepted papers. They have released a statement icml.cc/Conferences/... Yes, it’s unacceptable. So is using an LLM to review a paper. Peer review is so broken.

119833

Reposted by Ondrej Dusek

TBSkyen @tbskyen.com · 17/05/2025

#Eurovision is tonight, and here's a hilarious fun fact about it: Israel has started a massive offensive against civilian populations in Gaza with the explicit aim of conquering the entire territory and ethnically cleansing its population, and Eurovision has aggressively refused to give a shit.

1027361022

Reposted by Ondrej Dusek

Ben Collins @bencollins.bsky.social · 04/05/2025

It is a little weird to me countries aren’t more aggressively, formally trying to take advantage of the U.S. science brain drain. Once in a lifetime opportunity to buy low on Non-Dumbass Americans with PhDs who just wanna look into microscopes and quietly cure ass cancer as our country eats shit.

1220346254786

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 29/04/2025

The 👉Machine Learning Prague 2025👈 is happening right now! Today, @patuchen.bsky.social and @navitas.bsky.social presented their posters on text generation with LLMs. Also, don't miss @tuetschek.bsky.social's invited talk tomorrow at 11 a.m.

1115

Reposted by Ondrej Dusek

Tokenization Workshop (TokShop) @COLM2026 @tokshop.bsky.social · 15/04/2025

🚨 NEW WORKSHOP ALERT 🚨 We're thrilled to announce the first-ever Tokenization Workshop (TokShop) at #ICML2025 @icmlconf.bsky.social! 🎉 Submissions are open for work on tokenization across all areas of machine learning. 📅 Submission deadline: May 30, 2025 🔗 tokenization-workshop.github.io

tokenization-workshop.github.io

Tokenization Workshop @ ICML 2025

1247

Reposted by Ondrej Dusek

leon @leyawn.bsky.social · 24/04/2025

“is my calculator horny?“ our tech columnist asks. “i entered 5318008 into it and turned it upside down. what i saw surprised me”

156164663766

Reposted by Ondrej Dusek

Zdeněk Kasner @zdenekkasner.cz · 15/04/2025

How do LLMs compare to human crowdworkers in annotating text spans? 🧑🤖 And how can span annotation help us with evaluating texts? Find out in our new paper: llm-span-annotators.github.io Arxiv: arxiv.org/abs/2504.08697

llm-span-annotators.github.io

Large Language Models as Span Annotators

Website for the paper Large Language Models as Span Annotators

1207

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 09/04/2025

Participate in the 👉 CRAC 2025 Shared Task on Multilingual Coreference Resolution❗ ufal.mff.cuni.cz/corefud/crac25 If you have not already done so, register first. 👆 Then start discovering how words refer to each other in 1️⃣7️⃣ languages. This year includes a new ✨LLM✨ track 😮.

162

Reposted by Ondrej Dusek

Dare Obasanjo @carnage4life.bsky.social · 22/03/2025

It’s kind of quaint to think the big worry a few years ago was that AI chatbots would destroy humanity. We’re quite capable of doing that without their help.

614111

Reposted by Ondrej Dusek

Christian Wolf @chriswolfvision.bsky.social · 26/02/2025

The CVPR program chairs acted on the reviewing crisis in ML/CV conferences. This is a first! Papers of authors who acted as irresponsable reviewers have been desk rejected. 19 papers are affected in 2025. This has also been announced for ICCV 2025.

38011

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 20/02/2025

The 6th International Workshop on Designing Meaning Representation (#DMR2025) will be in Prague, Aug 4-5, right after #ACL2025 in Vienna! Submit your work on meaning representations: annotation, parsing, multilinguality, neuro-symbolic methods & more. Details: dmr2025.github.io/index

dmr2025.github.io

DMR 2025

The 6th International Workshop on Designing Meaning Representations

064

Reposted by Ondrej Dusek

Emma Strubell @strubell.bsky.social · 12/02/2025

Ok here's the theory so far for how the LLM-generated ARR submission experiment went down. Researchers generated papers, then created fake reviewer profiles by identifying existing papers from a real author st the set of papers would maximize expertise affinity score wrt the LLM-generated paper.

2277

Reposted by Ondrej Dusek

The Wheel Turns 〓〓 (Si) @unseenrealms.bsky.social · 23/02/2025

I’m doing the same. It’s important to do what we can. I’ve changed to Signal where possible, Proton mail, Startpage browser/search or , Qwant search on Firefox. This is a good site for European alternatives: european-alternatives.eu

european-alternatives.eu

Homepage | European Alternatives

We help you find European alternatives for digital service and products, like cloud services and SaaS products.

36824

Reposted by Ondrej Dusek

Maike Osborne @maosbot.bsky.social · 09/02/2025

The rise of ChatGPT isn't just about AI—it's about how cluttered the web has become. Compare these recipes: ChatGPT gives me ingredients instantly, while a cooking website shows me an auto-playing video and a prompt to log into Google so I can see an ad

The image shows a ChatGPT conversation where someone requested a recipe for pancakes using whole rolled oats without a blender, with measurements in grams. The recipe ingredients are listed as:

- 150g whole rolled oats
- 250ml milk (dairy or plant-based)
- 2 large eggs (approximately 100g)
- 30g melted butter (or neutral oil) + extra for cooking
- 20g sugar (optional, adjust to taste)
- 8g baking powder (about 2 tsp)
- 2g salt (about 1/2 tsp)
- 5ml vanilla extract (optional)

This image shows a webpage from "the kitchn" website featuring a recipe for "Easy Oatmeal Pancakes." The recipe appears to be highly rated with 5 stars from 64 reviews. The page shows a navigation path of RECIPES > BREAKFAST > PANCAKES, and there's a banner at the top advertising "Our Laziest, Most Delicious Dinners Ever (Ready in 30 Minutes or Less)." The image also shows part of a photo featuring what appears to be melted butter being whisked into a mixture. There's a Google sign-in prompt overlaying part of the page.

139712

Reposted by Ondrej Dusek

Institute of Formal and Applied Linguistics @ufal.mff.cuni.cz · 07/02/2025

We've been making the media rounds! 👉📺 @hajicjan.bsky.social talked about the new OpenEuroLLM project on Czech TV's Studio 6 www.ceskatelevize.cz/porady/10969... 👉📻 @tuetschek.bsky.social discussed #LLMs on Czech Radio radiozurnal.rozhlas.cz/proc-umela-i...).

ceskatelevize.cz

Čech povede evropský výzkum AI - 3. února 05:59 - Studio 6 | Česká televize

Jan Hajič

073