Happy to share that our work on evaluating tokenizers has been accepted to #ACL2025! This is one of two projects I worked on during my internship @Apple.

Postmortem: @UnslothAI Gradient Accumulation Report, @huggingface`transformers`, and You! First, what went wrong. A great visualization of this issue by @shxf0072is attached, essentially calculating the loss individually without taking into account when sequence lengths

[1/5] Are verbose #LLM translations skewing evaluation results? TLDR: Yes! Our recent work dives into the prevalence and impact of LLM verbosity in automatic and human evaluations. 📎 Paper: https://t.co/kPjVGuNGMp

I am thrilled to announce that a few months back, I joined Amazon in Madrid as an Applied Scientist II 🚀 In this role, I will be working on Cross-lingual Large Language Models and their intersection with search technologies. A big thank you to everyone who has supported me.

Delighted to share Generative Query Recommendations via LLMs. We show how LLMs enhance web search by improving queries in the absence of query logs. @enricopalumbo91, @adamianou, @ntonellotto, @fabreetseo. https://t.co/U5Cqezxps0 Work done prior to joining Amazon.

Happy to share that my internship work @Apple is finally available on @arxiv! Paper: https://t.co/docqNHrjEG Many thanks to my co-authors Thiago Fraga da Silva, Ernest Pusateri, Youssef Oualil and Markus Nussbaum.