Do you want to know what information LLMs prioritize in text synthesis tasks? Here's a short 🧵 about our new paper: an interpretable framework for salience analysis in LLMs. First of all, information salience is a fuzzy concept. So how can we even measure it?

Introducing ✨ Aya Vision ✨ - an open-weights model to connect our world through language and vision Aya Vision adds breakthrough multimodal capabilities to our state-of-the-art multilingual 8B and 32B models. 🌿

Are you studying, working in, or utilizing #forensics? We're looking for expert opinions in a short, anonymous survey! Message or email me with any questions or for the link! #Science #DNA #pathology

My team recently launched a best-in-class LLM specializing in English and Arabic. We just published a tech report explaining our methods. Check it out on arxiv: arxiv.org/abs/2503.14603

A year ago we released LBBP - a drop-in replacement of HumanEval that was more challenging and less leaked Internally we have been using the multilingual version of this for benchmarking, and as code is not only python we decided to release that as well huggingface.co/datasets/Coher…

𝐂𝐨𝐡𝐞𝐫𝐞 𝐄𝐦𝐛𝐞𝐝 𝐯𝟒 - 𝐒𝐭𝐚𝐭𝐞-𝐨𝐟-𝐭𝐡𝐞-𝐚𝐫𝐭 𝐭𝐞𝐱𝐭 & 𝐢𝐦𝐚𝐠𝐞 𝐫𝐞𝐭𝐫𝐢𝐞𝐯𝐚𝐥 Today we are releasing Embed v4, unlocking so many cool new features for retrieval. 🇺🇳 100+ languages 🖼️ Text & Image capabilities 📜 128k context length

AV's latest #BRIDGEseries convening brought together researchers, public officials, and industry experts to better understand the impact and prevalence of retail theft, and what can be done to effectively prevent it.

Command A, our state-of-the-art generative model, is now the highest-scoring generalist LLM on the Bird Bench leaderboard for SQL! It outperforms other systems that rely on extensive scaffolding to tackle these SQL benchmarks, and instead delivers these results out-of-the-box,

It's my first time area chairing for the ACLRollingReview May cycle! And it will also be the first time asking for availability of emergency reviewers 😅 If you (or somebody you know) has availability for reviews in the Resources and Languages track, I have two papers missing reviews.

Probably a good time to mention that I will be in Vienna attending ACL in two weeks. If you're tired of attending session after session, come and talk to me about LLM evaluation instead (I won't tell on you for skipping sessions🤫)! DMs are open if you want to set something up :)

Genuine question: how do people in tech with imposter syndrome survive the bay area?? It's bad enough elsewhere, but with the talent density there (plus supposedly being more openly bragging), it seems like death

👁️👁️ Cohere has a vision model now cohere.com/blog/command-a…

No secret to anyone who works with Pierre (and his team), but they are super cracked. Seeing (pun intended) this model come to life was amazing! Please try it out and let us know what you think 😎