Understanding and Mitigating Language Confusion in LLMs | Cohere | Research | TechiSeek

AddAdd your Listings

Research

Understanding and Mitigating Language Confusion in LLMs

We investigate a surprising limitation of LLMs: their inability to consistently generate text in a user's desired language.

Research Cohere

ResearchNo Need for Explanations: LLMs can implicitly learn from mistakes in-contextCohere ResearchFrom Tools to Teammates: Evaluating LLMs in Multi-Session Coding InteractionsCohere ResearchLanguage Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-ThoughtCohere ResearchAgents Explore but Agents Ignore: LLMs Lack Environmental CuriosityCohere ResearchProcedural Knowledge in Pretraining Drives Reasoning in Large Language ModelsCohere ResearchFishing for Magikarp: Automatically Detecting Under-trained Tokens in Large Language ModelsCohere BlogsThe value of llms.txt: Hype or real?Mintlify BlogsWhat is llms.txt? Breaking down the skepticismMintlify BlogsWhat is MCP and how to get startedMintlify Products & ServicesSupporting AI standards at GitBook: What OKF, MCP and llms.txt tell us about the future of docsGitbook BlogsCSS Intelligence: Speculating On The Future Of A Smarter LanguageSmashingmagazine BlogsSmashing Animations Part 3: SMIL’s Not Dead Baby, SMIL’s Not DeadSmashingmagazine BlogsMarch, April, and May 2025 updates 🚀Redocly BlogsWhat Is Natural Language Generation (NLG)?Zapier ResearchRewardBench 2: Advancing Reward Model EvaluationCohere ResearchThe Multilingual Divide and Its Impact on Global AI SafetyCohere ResearchHere's a Free Lunch: Sanitizing Backdoored Models with Model MergeCohere ResearchDiversify and Conquer: Diversity-Centric Data Selection with Iterative RefinementCohere ResearchSparkles: Unlocking Chats Across Multiple Images for Multimodal Instruction-Following ModelsCohere ResearchNexus: Specialization meets Adaptability for Efficiently Training Mixture of ExpertsCohere