I am very happy to present the multilingual expansion of REBEL at #ACL2023. This work has been a year-long effort to try and provide datasets and models that enable high-quality multilingual Relation Extraction on par with English.

Automatic knowledge graph construction can be slow and expensive. Also I find there's a lack of resources on how to build something principled (do you just stuff text into an LLM prompt?) That's why I love this blog by @tb_tomaz which not only outlines the step-by-uses Relik, a

Which layers memorise examples with permuted labels?🧐 That’s what we investigate in “Generalisation First, Memorisation Second? Memorisation Localisation for Natural Language Classification Tasks” presented in Findings poster session 2 @ 5.45PM today! #ACL2024 #NLProc (1/7)

I'm in Bangkok for #ACL2024NLP 🇹🇭 Looking forward to present our work "Dissecting Biases in Relation Extraction: A Cross-dataset analysis on People's Gender and Origin" @genderbiasnlp Joint work w/ @marcostranisci @PereLluisHC @RNavigli 📄 https://t.co/JrG9svJugT #NLProc 1/2

Check out our new work!

👀 Exciting News! 👀 Happy to announce our latest research paper, “Retrieve, Read and LinK: Fast and Accurate Entity Linking and Relation Extraction on an Academic Budget”, will be presented at #ACL2024! 🚀 Try it out! https://t.co/xJzHlNfEzA Thread below 👇

These headlines are 3 days apart. See, they can name the attacker without casting doubt.

ARR needs your help! We received 5800+ submissions for the June cycle, but with our current capacity, we can handle only half of these submissions. We can't start the review process without significant help. (1/2)

We are (finally) releasing the 🍷 FineWeb technical report! In it, we detail and explain every processing decision we took, and we also introduce our newest dataset: 📚 FineWeb-Edu, a (web only) subset of FW filtered for high educational content. Link: https://t.co/MRsc8Q5K9q

The new interpretability paper from Anthropic is totally based. Feels like analyzing an alien life form. If you only read one 90-min-read paper today, it has to be this one https://t.co/9yecogVqJf

GPU-Poor no more: super excited to officially release ZeroGPU in beta today. Congrats @victormustar & team for the release! In the past few months, the open-source AI community has been thriving. Not only Meta but also Apple, NVIDIA, Bytedance, Snowflake, Databricks, Microsoft,

Just interviewed by @alessiojacona @Agenzia_Ansa to talk about #LLMs and the @SapienzaNLP effort to create the first #Italian pre-trained #LLMs, the #Minerva family. Team: @edoardo_barba, @ConiaSimone, @perelluisHC, @AndrewWyn1, @RiccardoRicOrl & me

Join us today at #EACL2024 for our presentation entitled “CroCoAlign: A Cross-Lingual, Context-Aware and Fully-Neural Sentence Alignment System for Long Texts” in the Sentence-level Semantics track! #NLProc (Radisson Blu, Carlson Ballroom, 5th floor, Malta)

Exciting strides in text summarization with LLMs 🚀but verifying their factual accuracy is still an open challenge 🤔 We introduce FENICE, a factuality-oriented metric for summarization with a strong focus on interpretability🔍 https://t.co/jjEI6lbxzG #NLProc #LLMs #Factuality

2019 sueño vs 2024 realidad

🤯 Think adding nonsense to RAG systems is madness? Our new paper says otherwise! We found that including random documents boost accuracy by 30+%, challenging old paradigms and showing the complexity of integrating retrieval w/ language generation. #RAGSystems #surprisingresults

New Anthropic Paper: Sleeper Agents. We trained LLMs to act secretly malicious. We found that, despite our best efforts at alignment training, deception still slipped through. https://t.co/mIl4aStR1F

I've written my first #Medium story! I delve into the longstanding challenge of #AGI, and provide an high-level overview of the current #NLP landscape, highlighting both the milestones achieved and the persisting challenges! 🧗🏻 #ChatGPT #LLMs #AI https://t.co/yHloZLhZaD📑

It's been extremely disappointing compared to the experience in ACL venues. In our paper not a single reviewer has acknowledged our answers. Did reviewers receive any reminders to engage @iclr_conf?

I am seriously concerned about the quality and tone of many @iclr_conf reviews I have read in both my reviewer and author batches. I see - Reviews likely written by a language model - Dismissive remarks - Judgemental comments - Impolite tone - ... https://t.co/bbPGdxkZe5

If you are looking for a PhD position on knowledge graphs and machine learning, this might be an opportunity for you 👇