🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Felix is directly responsible for me getting into NLP by responding to a cold email from a clueless maths grad and giving me a summer project. RIP Felix, every subsequent opportunity I’ve had has been downstream of that one small act.

I’m so excited to share something we’ve been working on for awhile: North is cohere’s AI workspace for enterprises. Today we’re releasing the platform for early access!

As someone who briefly touched the transcendent (applied for PhDs in axiomatic set theory) this resonates strongly with me. I think that ‘genius’ is as much the effort of self-cultivation as it is birth, and that the route to mastery in any subject is surprisingly similar.

cohere is growing - We're hiring MLE's to build North in our London, New York and Toronto offices. We also support remote working, 'cos RTO mandates aren't our thing. Job Spec is below in the 🧵, with more info on North available here; cohere.com/north

📷 Excited to share our new paper: "Rope to Nope and Back Again: A New Hybrid Attention Strategy" where we propose a novel architecture that outperforms RoPE-NTK-based approaches with full attention span. (1/8)

once again the function-space view of neural networks leads to actionable insights. gaussian processes should (once again) be required knowledge for ML.

I'm hiring performance engineers for the pre-training team at Cohere. If you enjoy writing efficient kernels, hardware-aligned architecture design and optimisations, do reach out! Check out the live job posting here: jobs.ashbyhq.com/cohere/d42f5fd…

Training on a little 🤏 formal language BEFORE natural language can make pretraining more efficient! How and why does this work? The answer lies…Between Circuits and Chomsky. 🧵1/6👇

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. 🌿

we have a new model, it's pretty good and we like it, we think you'll like it too. (as an aside this is the first model i contributed to at cohere!)

thinking about tree lstms again

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

I'm excited to the tech report for our @Cohere Cohere For AI Command A and Command R7B models. We highlight our novel approach to model training including the use of self-refinement algorithms and model merging techniques at scale. Command A is an efficient, agent-optimised

in case you are looking for the best model for COBOL, we might be able to help you here... cohere.com/research/paper…

Wanna check how well a model can share knowledge between languages? Of course you do! 🤩 But can you do it without access to the model’s weights? Now you can with ECLeKTic 🤯

the acl paper template will long outlive acl as an attractive destination for NLP papers

If you are based in Zurich (or anywhere rly) and write code for ML accelerators (including cuda/rocm) HMU

rewinding model loss spikes is like saving before a difficult boss and reloading if you start losing

How can we make language models more flexible to adapt to new languages after pretraining? 🌏 🧠 Our latest work investigates whether a tokenizer trained on more languages than the pretraining target can improve language plasticity without compromising pretraining performance.

🚨New pretraining paper on multilingual tokenizers 🚨 Super excited to share my work with Cohere Labs: One Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual Tokenizers