Thrilled to share our new paper: "Can LLM be a Personalized Judge?" We investigate the reliability of LLMs in judging user preferences based on personas and propose improvements using verbal uncertainty estimation to enhance accuracy. 🎭👨‍⚖️ 📄 Paper:

I've developed a handy way of streamlining our @emnlpmeeting reviewer criticisms on social media

Americans: I use miles and pounds Europeans: I use kilometres and kilograms Canadians: [snorting a line of assorted measuring systems] I'm 5'3, I weigh 150lbs, horses weigh 1000kgs, my house is an hour away and I drive 80 km/h to get there, I need a cup of flour and 1L of milk

We present LongForm📜, optimizing instruction tuning for long text generation with corpus extraction. LongForm models outperform baselines such as FLAN-T5 and Alpaca. 📄Paper (w/@timo_schick @annalkorhonen @HinrichSchuetze): https://t.co/rWJZn7DpcD 💾Repo: https://t.co/aCJkvaMpA7

We are hiring several short-term student assistants for a chatbot annotation study, conducted by @PetrakDominic and @thy2512 in the context of @SERMAS_eu (Socially-acceptable Extended Reality Models and Systems). #XR Interested? More infos here ➡️

Here we are! ✨Multi3NLU++: A Multilingual, Multi-Intent, Multi-Domain Dataset for Dialogue NLU✨ Multi3NLU++ contains user input-based examples in 2 domains, 4 languages and labeled with 62 intent types and 17 slots. 📄 https://t.co/MbMu0HuFbC 👩‍💻

my hot take is that you should not advertise if ur paper was accepted but instead what ideas it has my even hotter take is that if you think that the 'number' of papers that you/your lab have submitted vs. accepted is something to be proud of then you should quit science asap

The Big Lebowski as a metaphor for Machine Learning academia, a thread. Walter, a PhD student who bets all his money on a single hypothesis, doesn't even think that there might be an alternative explanation, feels surprised when the hypothesis is refuted.

At Google, female research scientist women were blocked when they tried to hire other female research scientist women. The reason given, on multiple occasions, was that we were trying to hire "our friends". Shall we unpack this with respect to how sexist discrimination works?

Fast, flexible cultural transmission underpins human intelligence. Our team trains an AI capable of real-time cultural transmission in previously unseen navigation tasks. The agent follows expert demos & reproduces them reliably after the expert leaves: https://t.co/AGkDnb2bgh 1/

Step 3: Tech twitter & especially #AIethics twitter have a field day, including such gems as: https://t.co/IUDvK24dpg

Our paper "Annotation Curricula to Implicitly Train Non-Expert Annotators" was accepted at Computational Linguistics! 🥳 We sincerely thank our anonymous reviewers and editor for their constructive reviews and patience throughout multiple rounds! https://t.co/V9tTy6q8vQ #NLProc

🚨 We’re releasing QuALITY, a benchmark for reading comprehension with long texts! 🚨 Yes, the acronym is a little tone-deaf, but this is almost certainly the best benchmark or dataset release from my group so far. (🧵)

This.

Types of ML / NLP Papers

Drug dealer shockingly bad at getting people drugs https://t.co/f4RMa3IY4T #FordFailedThePeople #FordfailedOntario

Dear authors working on your rebuttals: this is your reminder that the below issues should NOT be cited as a major weakness of a paper. If you received such a review, please flag the issue both in response to the reviewer and in the “response to chairs” box. [1/2]