🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Serious question: What does an NLP Ph.D student work on nowadays with the presence of closed source GPT models that beat anything you can do in standard academic lab? Sam Bowman Sasha Rush Christopher Manning Mark Dredze Christopher Potts

GPT-4 and its ilk are awesome for rapid prototyping and one-offs, but at the end of the day, enterprises will deploy far smaller distilled models in production. Here's my contrarian take -

A survey of LLMs with a practical guide and evolutionary tree. Number of LLMs from Meta = 7 Number of open source LLMs from Meta = 7 The architecture nomenclature for LLMs is somewhat confusing and unfortunate. What's called "encoder only" actually has an encoder and a decoder

Today we are announcing a major breakthrough in the Vesuvius Challenge: we have read the first word from an unopened Herculaneum scroll. The word is "πορφυρας" which means "purple dye" or "cloths of purple." scrollprize.org/firstletters Congratulations to 21yo computer science

Which dense retriever should you use for zero-shot search? Ekaterina Khramstova is giving us the answer at the #SIGIR_AP2023 #MelbpurneSatellite

A version of Llama-2 fine-tuned for medicine from EPFL.

Also! Xinyu Mao, Bevan Koopman, and Guido Zuccon will present their reproducibility study on my paper with Sean MacAvaney “Goldilocks: Just-Right Tuning of BERT for TAR” on Wednesday afternoon. I will be there to figure out whether I messed up my experiments!

#sigirap2024 call for papers, featuring the SIGIR Revise-and-Resubmit paper type, is available now! sigir-ap.org/sigir-ap-2024/…

Very happy to share that our work "Dense Retrieval with Continuous Explicit Feedback for Systematic Review Screening Prioritisation" w/ Shengyao Zhuang, Bevan Koopman, Guido Zuccon has been accepted at SIGIR 2025 #SIGIR2024 as a short paper. Excited for my first #SIGIR experience!

.Xinyu Mao now presenting his work on reproducing the work of Eugene Yang titled "A Reproducibility Study of Goldilocks: Just-Right Tuning of BERT for TAR" at #ecir2024

Joining the merry #sigir2024-rejection gang: Happy to share our paper on the Set-Encoder is now on arXiv arxiv.org/abs/2404.06912. It features permutation-invariant listwise re-ranking with an ELECTRA-base model that is as effective as RankGPT4. Code and data available 😉

Methodological insights into ChatGPT’s screening performance in systematic reviews bmcmedresmethodol.biomedcentral.com/articles/10.11…

We have a PhD scholarship and 2 postdoc openings in the ielab team to work on Information Retrieval and LLMs. Come and join us! 🧵

Now Shengyao Zhuang presenting our setwise LLM ranker at #sigir2024

.Niklas Deckers now presenting "Evaluating Generative Ad Hoc Information Retrieval" at #sigir2024

Our Llama 3.1 405B is now openly available! After a year of dedicated effort, from project planning to launch reviews, we are thrilled to open-source the Llama 3 herd of models and share our findings through the paper: 🔹Llama 3.1 405B, continuously trained with a 128K context

Our new paper, which studies the vulnerability of document screenshot retrievers like DSE and ColPali to pixel poisoning attacks, is now available on Arxiv! arxiv.org/pdf/2501.16902 This work was done with Ekaterina Khramtsova, Xueguang Ma, Bevan Koopman, Jimmy Lin, Guido Zuccon.

🎉 Excited to share the list of our ielab papers accepted at #SIGIR2025 ACM SIGIR! Topics include efficiency in IR, domain-specific IR (SysRev, podcast), screenshot retrievers, hallucination detection, and more. See you in Padua 🇮🇹! Details coming soon, stay tuned!

Looking forward to presenting our demo tool AiReview at #SIGIR2025 in Padua—🔍 showcasing how LLMs can accelerate systematic review screening! BTW, our preceding dense retrieval-based tool, DenseReviewer, will be at #ECIR2025 demo—check it out here: arxiv.org/abs/2502.03400 🚀

.Shengyao Zhuang Xueguang Ma and I are looking forward to tomorrow's "R2LLMs: Retrieval and Ranking with LLMs" tutorial at #sigir2025 "Light" agenda at ielab.io/tutorials/r2ll… -> slides and other material will be available here too soon SIGIR 2025