Thrilled to share our work on IndicMMLU-Pro, a benchmark for evaluating LLMs in 9 Indic languages!
Working on a topic close to my heart has been a rewarding experience!!Quality benchmarks for Indic languages are crucial for inclusive AI progress.
https://t.co/2hKzfK1j6n
New year, new paper! 💡
Mirror, mirror on the wall, can #LLMs judge LLMs after all? Our latest research puts LLMs to the test, evaluating their ability to judge LLM-generated text analytics outputs.
Read more: https://t.co/Zd0iAFZY6x
@i_amanchadha@AmazonScience
LLM-augmented clustering enables Amazon’s Qualitative Insights Tool (QualIT) to outperform other topic-modeling methods when extracting information from employee surveys, product feedback channels, or other unstructured text sources: https://t.co/F91Ovkehph #LLMs
Check out this new chapter by Xin Luo, Chunping Fan, and Fumihiko Tochinai and me – open access, please share widely!
“Two Criticisms of Engineering Ethics Assessment: The Importance of Behaviors and Culture”
What’s wrong with approaches to technology ethics assessment?
Shoutout to the incredible cast of @CursedChildNYC! Big shoutout to the Malfoy father-son duo and Special thanks to the cast for the post-show love, patiently posing and signing playbills despite the bitter cold - you truly warmed our hearts and made it magical. 🥰❤️🩹
Just spent the evening under the Hogwarts spell! @CursedChildNYC left me laughing, crying (Snape has my heart, always), and utterly enchanted.
Each act was phenomenal!
Thankful for @TodayTix and their #FridayForty lottery for making my Sunday magical. 🪄
@1810Kali It was incredible! The Bengali club USA did a phenomenal job! And so many artists from across Bangladesh and India, right there at Times Square!