Thrilled to have contributed (although in a tiny way!) to this amazing release alongside such an brilliant team.

Everyone is on their way to Vancouver for NeurIPS Conference 2024! ✈️ We are looking forward to connecting with everyone! 🤗 Here is where you can find us:

Started the morning off with the BEST pain au chocolat and views in Vancouver for #NeurIPS2024 🥐🏔️ Presenting this work at the Safe Generative AI workshop on Sunday, December 15! I’ll also be at the Cohere booth later today to chat all things model merging, safety and more! 💫

Time to prove that I'm not all about that LLM bandwagon -- presenting work w/ Yash Chandak Yannis Flet-Berliac, Christina, Ani, and Emma Brunskill on Offline RL policy evaluation. Friday Dec 13: 11am-2 pm Poster Location: West Ballroom A-D Stanford AI Lab Stanford HAI

📬 This is what we've been cooking the last few months - check out our new 7B model! Smaller and faster, it beats Gemma-2 9B, Ministral 8B and Llama-3.1 8B on the OpenLLM leaderboard. It has best-in-class RAG, tool use, agents + enhanced efficiency in reasoning, maths and code.

Trying out Command R7B in VSCode, and the model performs brilliantly! 👏 The latest model from cohere Command family shows excellent performance working with code inside VSCode, using CodeGPT to integrate the model. Congrats to the Cohere team! 🥳 If you want to use Cohere's

Command R7B is open-weights. ✨Try it on the Cohere For AI Hugging Face space or download it directly here: huggingface.co/CohereForAI/c4…

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!

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’re excited to introduce our newest state-of-the-art model: Command A! Command A provides enterprises maximum performance across agentic tasks with minimal compute requirements.

Excited to share Command A! It outperforms GPT-4o and DeepSeek-V3 with unmatched efficiency. It excels in business, STEM, and coding tasks while being 1.75x faster than GPT-4o and 2.4x faster than DeepSeek-V3. Glad to have contributed to its post-training with a fabulous team!

Following the open-weight release of Command A and Command R7B models, we're excited to have collaborated with @Cohere colleagues on a tech report highlighting our novel approach to model training, including self-refinement algorithms and model merging techniques at scale.

Excited to finally share that CoPG — the RL method I co-authored with Nathan Grinsztajn and amazing colleagues — was used throughout the post-training (offline & online learning) of cohere’s new Command models! 🖊️ Tech report: cohere.com/research/paper… 🤖 CoPG: arxiv.org/abs/2406.19185

Would you like to use Q-learning for LLM fine-tuning? Check out our new preprint where we interpret Q-functions as logits of the LLM: arxiv.org/abs/2505.11081 ✨ Work done with my based colleagues cohere

Checkout our new vision model! Now available on our platform and coming soon to Hugging Face for research use.

Our new (and first!) reasoning model. > feel the egi