NLP & ML research @cohereforai.bsky.social 🇨🇦https://bsky.app/profile/juliakreutzer.bsky.socialhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3mpjolzsg7s2r🤨"But why linguistics" is the most common question when talking about linguistic reasoning benchmarks. Last year we organized a shared task at WMT...and no one participated 🤣 🤯Let me change your mind why this is one of the most challenging, focused and best reasoning benchmarks right now. [contains quote post or other embedded content]30 Jun 2026 18:57 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3mpjolzsg7s2rhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3mpjoh7yuxk2r🔥Possibly the most fun and underrated AI challenge this summer: Compete on unseen linguistic reasoning problems and present your solutions to the expert jury in a month! [contains quote post or other embedded content]30 Jun 2026 18:54 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3mpjoh7yuxk2rhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3mgvlal6hfs2oJoin us tomorrow for a discussion around culture in AI! 🌱 [contains quote post or other embedded content]12 Mar 2026 23:42 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3mgvlal6hfs2ohttps://bsky.app/profile/juliakreutzer.bsky.social/post/3mg6jkhbk522y💭We need more research that focuses on aspects beyond accuracy, especially in multilingual AI. 👉Help us explore the importance of culture in building and testing AI, with a few minutes of your time. Happy to have a chat as well with anyone who's interested in that space! [contains quote post or other embedded content]03 Mar 2026 19:40 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3mg6jkhbk522yhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3mf3z4lyjw226🌱Very proud of our team's latest release 😊 meet Tiny Aya, a massively multilingual model with 3.35B parameters. Tech report here: https://github.com/Cohere-Labs/tiny-aya-tech-report/blob/main/tiny_aya_tech_report.pdf [contains quote post or other embedded content]18 Feb 2026 02:16 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3mf3z4lyjw226https://bsky.app/profile/juliakreutzer.bsky.social/post/3m6zb4cxe6c2sAt #Neurips2025 this week with @cohereforai.bsky.social 🤩 This is what brought me into research: the ✨Network Effect ✨ Let's build the next breakthrough together! #LabLegends02 Dec 2025 15:29 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3m6zb4cxe6c2shttps://bsky.app/profile/juliakreutzer.bsky.social/post/3m2tny2u6kc2iLet's do the venue justice. Very excited for today's multilingual workshops at #COLM2025 💙 [contains quote post or other embedded content]10 Oct 2025 12:23 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3m2tny2u6kc2ihttps://bsky.app/profile/juliakreutzer.bsky.social/post/3m2sbzfpeq224Looking forward to tomorrow's #COLM2025 workshop on multilingual data quality! 🤩 [contains quote post or other embedded content]09 Oct 2025 23:16 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3m2sbzfpeq224https://bsky.app/profile/juliakreutzer.bsky.social/post/3m2omoagdpc2jReady for our poster today at #COLM2025! 💭This paper has had an interesting journey, come find out and discuss with us! @swetaagrawal.bsky.social @kocmitom.bsky.social Side note: being a parent in research does have its perks, poster transportation solved ✅ [contains quote post or other embedded content]08 Oct 2025 12:16 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3m2omoagdpc2jhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3m27q7bzpx22d💡A collaborative➕diverse team is key. In real life as in the LLM world 💪🦾 Check out our latest work that builds on this insight. 👇 [contains quote post or other embedded content]02 Oct 2025 14:10 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3m27q7bzpx22dhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3lsjqdlg42k2y🍋 Squeezing the most of few samples - check out our LLMonade recipe for few-sample test-time scaling in multitask environments. Turns out that standard methods miss out on gains on non-English languages. We propose more robust alternatives. Very proud of this work that our scholar Ammar led! 🚀 [contains quote post or other embedded content]26 Jun 2025 18:17 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3lsjqdlg42k2yhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3lqrq4egy2c2l🚨LLM safety research needs to be at least as multilingual as our models. What's the current stage and how to progress from here? This work led by @yongzx.bsky.social has answers! 👇 [contains quote post or other embedded content]04 Jun 2025 11:44 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3lqrq4egy2c2lhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3lqaj73cu4s23🚧No LLM safety without multilingual safety - what is missing to closing the language gap? And where does this gap actually originate from? Answers 👇 [contains quote post or other embedded content]28 May 2025 15:25 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3lqaj73cu4s23https://bsky.app/profile/juliakreutzer.bsky.social/post/3lor7otze3c2hMultilingual 🤝reasoning 🤝 test-time scaling 🔥🔥🔥 New preprint! @yongzx.bsky.social has all the details 👇 [contains quote post or other embedded content]09 May 2025 20:00 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3lor7otze3c2hhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3lmzqkyengk2y🤓MT eyes on multilingual LLM benchmarks 👉 Here's a bunch of simple techniques that we could adopt easily, and in total get a much richer understanding of where we are with multilingual LLMs. 🍬Bonus question: how can we spur research on evaluation of evaluations? [contains quote post or other embedded content]17 Apr 2025 18:33 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3lmzqkyengk2yhttps://bsky.app/profile/juliakreutzer.bsky.social/post/3lmyx2tsn3c2z📖New preprint with Eleftheria Briakou @swetaagrawal.bsky.social @mziizm.bsky.social @kocmitom.bsky.social! arxiv.org/abs/2504.11829 🌍It reflects experiences from my personal research journey: coming from MT into multilingual LLM research I missed reliable evaluations and evaluation research…17 Apr 2025 10:56 +0000at://did:plc:rmosvsit2iwn6hhztsoojn7k/app.bsky.feed.post/3lmyx2tsn3c2z