Language Technologies Institute | CMU

We're excited to announce the launch of MMU-RAG at #NeurIPS2025! A new competition from the LTI & Amazon AGI focused on real-world Retrieval-Augmented Generation (RAG) with live user feedback via RAG-Arena.

Read more: lti.cmu.edu/news-and-eve...

Congrats (and well done!) to LTI PhD student Shaily Bhatt on her plenary lightning talk at #IC2S2

Hand gestures are a major mode of human communication, but they don't always translate well across cultures. New research from @akhilayerukola.bsky.social, @maartensap.bsky.social and others is aimed at giving AI systems a hand with overcoming cultural biases:
lti.cmu.edu/news-and-eve...

Hand gestures are a major mode of human communication, but they don't always translate well across cultures. New research from @akhilayerukola.bsky.social, @maartensap.bsky.social and others is aimed at giving AI systems a hand with overcoming cultural biases:
lti.cmu.edu/news-and-eve...

New research from LTI, UMich, & Allen Institute for AI: LLMs don’t just hallucinate – sometimes, they lie. When truthfulness clashes with utility (pleasing users, boosting brands), models often mislead. @nlpxuhui.bsky.social and @maartensap.bsky.social discuss the paper:
lti.cmu.edu/news-and-eve...

Analysis | 6 tips to avoid using AI chatbots all wrong Chatbots are an embarrassing mistake waiting to happen. Here’s what not to do.

Niloofar Mireshghallah, incoming @cylab.bsky.social ky.social‬, @cmu-epp.bsky.social, and @ltiatcmu.bsky.social professor, shares tips with @shiraovide.bsky.social and @heatherkelly.bsky.social of @washingtonpost.com for protecting #privacy while using #chatbots.

What if AI played the role of your sassy gay bestie 🏳️‍🌈 or AAVE-speaking friend 👋🏾?

You: “Can you plan a trip?”
🤖 AI: “Yasss queen! let’s werk this babe✨💅”

LLMs can talk like us, but it shapes how we trust, rely on & relate to them 🧵

📣 our #FAccT2025 paper: bit.ly/3HJ6rWI

[1/9]

Congratulations to LTI Assistant Prof @ybisk.me on his Amazon Research Award!

lti.cmu.edu/news-and-eve...

Our MSAII and MCDS programs were both recently ranked among the nations best by TechGuide, with the MSAII program being named the number one program among MS programs in AI. Congrats to all the faculty, students and staff that make us who we are! lti.cmu.edu/news-and-eve...

We had a blast at our Commencement Reception Friday night. Congrats again to all of our 2025 grads – go out and make us (even more) proud!

Notice our new look? We're thrilled to unveil our new logo – representing our vision, values, and the future ahead. Stay tuned for more!

“Go forth and innovate. Disrupt wisely. And carry your LTI pedigree with pride.”

Read LTI Director Mona Diab’s message to our 2025 graduating class:

www.lti.cmu.edu/news-and-eve...

Congratulations to all of the MIIS, MSAII, MCDS and MLT graduates receiving their diplomas today! 🎓🎉 #tartanproud

For 125 years, @cmu.edu has been achieving what was never thought possible. Celebrate #CMU125 and find out what you can expect during this milestone anniversary year: www.cmu.edu/125?utm_sour...

Excited to announce our #NAACL2025 Oral paper! 🎉✨

We carried out the largest systematic study so far to map the links between upstream choices, intrinsic bias, and downstream zero-shot performance across 131 CLIP Vision-language encoders, 26 datasets, and 55 architectures!

If you're at NAACL this week (or just want to keep track), I have a feed for you: bsky.app/profile/did:...

Currently pulling everyone that mentions NAACL, posts a link from the ACL Anthology, or has NAACL in their username. Happy conferencing!

Ever trusted a metric that works great on average, only for it to fail in your specific use case?

In our #NAACL2025 paper (w/ @841io.bsky.social), we show why global evaluations are not enough and why context matters more than you think.

📄 aclanthology.org/2025.finding...
#NLP #Evaluation

(🧵1/9)

Can self-supervised models 🤖 understand allophony 🗣? Excited to share my new #NAACL2025 paper: Leveraging Allophony in Self-Supervised Speech Models for Atypical Pronunciation Assessment arxiv.org/abs/2502.07029 (1/n)

🚀 Excited to share a new interp+agents paper: 🐭🐱 MICE for CATs: Model-Internal Confidence Estimation for Calibrating Agents with Tools appearing at #NAACL2025

This was work done @msftresearch.bsky.social last summer with Jason Eisner, Justin Svegliato, Ben Van Durme, Yu Su, and Sam Thomson

1/🧵

When interacting with ChatGPT, have you wondered if they would ever "lie" to you? We found that under pressure, LLMs often choose deception. Our new #NAACL2025 paper, "AI-LIEDAR ," reveals models were truthful less than 50% of the time when faced with utility-truthfulness conflicts! 🤯 1/

In-Context Learning with Long-Context Models: An In-Depth Exploration Amanda Bertsch, Maor Ivgi, Emily Xiao, Uri Alon, Jonathan Berant, Matthew R. Gormley, Graham Neubig. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Co...

and @abertsch.bsky.social, @emilyxiao.bsky.social, @gneubig.bsky.social and Matt Gormley won the SAC Award for Language Modeling for their paper "In-Context Learning with Long-Context Models: An In-Depth Exploration"
aclanthology.org/2025.naacl-l...

REL-A.I.: An Interaction-Centered Approach To Measuring Human-LM Reliance Kaitlyn Zhou, Jena D. Hwang, Xiang Ren, Nouha Dziri, Dan Jurafsky, Maarten Sap. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguisti...

@maartensap.bsky.social won Best Paper Runner-Up for his paper "REL-A.I.: An Interaction-Centered Approach To Measuring Human-LM Reliance" aclanthology.org/2025.naacl-l...

@seungonekim.bsky.social, @gneubig.bsky.social and Sean Welleck won Best Paper for their paper "The BiGGen Bench: A Principled Benchmark for Fine-grained Evaluation of Language Models with Language Models" aclanthology.org/2025.naacl-l...

Congrats to the many LTI researchers with accepted papers at #NAACL2025, and to our award winners!

Diab: “This opportunity to contribute to the Computing Research Association's mission comes at a pivotal moment for our field—an inflection point where the future of computing must be shaped with both ambition and responsibility.”

ICYMI: LTI Director and Professor Mona Diab was named to the board of directors of the Computing Research Association: www.lti.cs.cmu.edu/news-and-eve...

YouTube video by Language Technologies Institute at Carnegie Mellon (LTI at CMU) 4.18.25 LTI Colloquium Valentin Hofmann

Next was a fantastic talk by @valentinhofmann.bsky.social on probing covert racism in LLMs at the @ltiatcmu.bsky.social. One can imagine where this goes. Highly recommend www.youtube.com/watch?v=_1Ej... (5/11)

Congrats to the Purpl3Pwn3rs and Team RedTWIZ! Both teams feature LTI students, and both are finalists in the inaugural Amazon Nova AI Challenge. Read about it here:

www.lti.cmu.edu/news-and-eve...

A warm welcome to all of our Open House attendees this week – enjoy Pittsburgh, future Tartans!

Fact Check: BBC journalist’s hot mic moment on Trump-Zelenskiy clash is fake A recording of the BBC’s Ukraine correspondent seemingly caught by a hot mic making a crude joke about the White House clash between U.S. President Donald Trump and Ukrainian President Volodymyr Zelenskiy is not authentic, according to a BBC spokesperson and two digital media experts.

Rita Singh, faculty in CMU's Language Technologies Institute provided her expert analysis on a recording of a BBC correspondent, apparently caught making a crude joke about the clash between U.S. President Trump and Ukrainian President Volodymyr Zelenskiy.