The WMT'24 edition of the QE Shared Task it out! www2.statmt.org/wmt24/qe-task.… Focus? automatic methods for estimating the quality of NMT output at run-time. **NEW** this year => a subtask to provide explanations for the detected errors, as well as, suggest corrections! [1/n]

Mexico City is amazing! I am at NAACl24 🇲🇽. I will present our poster LLMRefine (using fine grained feedback to significantly improve llm’s generation quality) at 11AM of Wednesday (Mexican local time) at session 7. Please come by and talk to us if you are interested!

The Google Translate team had an amazing experience in Mexico City at #NAACL. We connected with many old and new friends and presented three papers! A big thank you to everyone who contributed to making this conference such a great experience. It was wonderful to see so many

The playlist starts in West Africa and wanders West-to-East until it hits Brazil. It’s composed of songs supplied by native speakers, which tend to be bangers. youtube.com/playlist?list=…

📢WMT24 Quality Estimation Shared Task is ON! 🚨 Submit your QE systems to predict translation quality, detect error spans or even correct the translations! Show us how you can improve referenceless MT evaluation! #WMT #MachineTranslation Submission deadline: July 31 (AoE)

We released WMT preliminary ranking of systems and LLMs in the General MT task. The official human evaluation and detailed insights will be ready in WMT conference findings. arxiv.org/abs/2407.19884

I am excited to attend ACL2024 to present our proud paper "Pride and Prejudice: LLM Amplifies Self-Bias in Self-Refinement" (We received scores of 5, 4.5, 4 and a meta-review of 5🤲). I am looking for full-time industry positions. Here is a list of my first-author publications

I’m super thrilled to have won the AMTA Best Thesis Award!! A huge thanks to the AMTA organizers for this recognition ☺️ See you all in Chicago amtaweb.org

🥳 LLMs are changing the game, even for datasets! NewsPaLM, a publicly released LLM-generated dataset, outperforms larger web-crawled corpora for MT. It includes sentence & paragraph-level, MBR-decoded data. See paper for more, incl. LLM self-distillation. arxiv.org/abs/2408.06537