Haritz Puerto

1/ A year ago, we introduced "Firewalls to Secure Dynamic LLM Agentic Networks" — the first architecture for securing agent-to-agent communication. Back then, autonomous agent networks felt speculative. Not anymore. We release a major update! 🧵👇 🔗 arxiv.org/abs/2502.01822

I am excited to announce that I have joined @sahar_abdelnabi's group at @ELLISInst_Tue and @MPI_IS as a postdoc! Looking forward to exciting projects in AI Safety!

@josephimperial_ Strange a desk rejection for such a small and subjective thing... In the past I had to fight really hard to desk reject clear cases of plagiarism (in the end the paper went through the normal cycle and reject, but not desk rejected) These decisions seem so random...

Excited to share that my paper was desk rejected for referencing a table in the appendix #ACL2026

Some thoughts on #NeurIPS2025, by someone who mostly attended vision conferences (me) - there were so many company booths about finance. There was even Wells Fargo ffs - way more evening events than at CVPR - there were many non-technical people (very rare at CVPR) [1/3]

How to get citations?

@Laz4rz @MistralAI It would make a lot of sense! Tübingen AI is growing a lot and the latent is huge. Cooperation with academia (mpi, Ellis, university) would efficient and effective and as you say it would be easier to hire that talent

so @MistralAI just opened in Zurich and Lausanne lol With Paris, Zurich (ETH), Lausanne (EPFL), Warsaw (UW), they're sucking like 70% of EU talent, only Tubingen missing

Wondering how to attend an ML conference the right way? ahead of NeurIPS 2025 (30k attendees!) here are ten pro tips: 1. Your main goals: (i) meet people (ii) regain excitement about work (iii) learn things – in that order. 2. Make a list of papers you like and seek them out at poster sessions. Try to talk to the authors– you can learn much more from them than from a PDF. 3. Pick one workshop and one tutorial that sounds most interesting. Skip the rest. 4. Cold email people you want to meet but haven't. Check Twitter and the accepted papers list. PhD students are especially responsive. 5. Practice a concise pitch of unpublished research you're working on for "what are you interested in rn?". Focus on big unanswered questions and exciting new directions, *not* papers. 6. Skip the orals. Posters are a higher-bandwidth, more engaging, more invigorating. Orals are a good time to go for a walk or talk in the hallway. 7. for the love of god, do NOT work on other research in your hotel room. Save mental bandwidth for the conference. (This may seem obvious; you'd be surprised.) 8. Talk to people outside your area. There are many smart people working on niches <10 people understand. Learn about one or two that won't help your own work. 9. Attend one social each night. Don't overthink it or get caught up in status games. They're all fun. 10. Take breaks. You can't go to everything, and conferences consume more energy than a normal workweek. hope this helps, and sad i'm not attending neurips, have fun :)

#NeurIPS2025 how does visibility work in AI search? Is SEO dead? Come to poster 1802 to find out! TLDR: no

#NeurIPS2025 starting! Crazy how large it is!

I will be presenting C-SEO Bench, does conversational SEO work? at #NeurIPS2025 🇺🇸 on Wednesday at 4:30PM, Exhibit Hall C,D,E #1802. Come if you want to see that SEO remains vital for AI search🔎!

Traveling to #NeurIPS2025 to present our poster about visibility in AI search. Is #seo dead? Not so fast! 👇🧵

Come to our #EMNLP2025 poster on Friday 7th at 10:30-12:00 in Hall C3! You will learn that reasoning traces in LRMs are not that safe and private as you may think 🫗

We are still looking forward for your submissions to the ELLIS UnConference LLM Safety and Security Workshop! @EurIPSConf @ELLISforEurope

Dr.LLM: Dynamic Layer Routing in LLMs Neat technique to reduce computation in LLMs while improving accuracy. Routers increase accuracy while reducing layers by roughly 3 to 11 per query. My notes below:

@jparapar @anxopvila @sepln congrats @anxopvila !! 🎉

Couldn’t be prouder! 🎉 @anxopvila has received @sepln’s Best NLP PhD Thesis Award, a huge honor and well-deserved recognition of his work. 👏🌟

Work with an amazing team! @framart1 @tommasogreen and @coallaoh from @parameterlab @oodgnas from @NAVER_AI_Lab

Excited to announce that my Conversational SEO paper got accepted at #neurips25 D&B!! 🎉🥳 Do you want to know how do to SEO on conversational search engines such as Perplexity? Take a look to our paper and thread 👇 or come to our poster in 🇺🇸!

Friends in Europe working in industry on Generative AI, can you help my friend with this survey. Participation is compensated with $40 / €36 Amazon gift card!

Excited to be part of this work! Congrats all the authors! 🎉 #EMNLP2025 Check the thread from @omarsar0 if you want to know more about the privacy risks introduced by the reasoning traces in reasoning models 🤓 x.com/omarsar0/statu…

I’ll be presenting today at 11:00 in hall x5 booth 209 #ACL2025NLP come and let’s talk about how to train with CoTs!