mlsanityData to 7 Oct 2026

Aya Expanse: Combining Research Breakthroughs for a New Multilingual Frontier

arXiv 2412.04261 ↗First posted 5 Dec 2024cs.CLcited by 121

Not on the map. Why?

Cited by 22 in LLM reasoning, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

We introduce the Aya Expanse model family, a new generation of 8B and 32B parameter multilingual language models, aiming to address the critical challenge of developing highly performant multilingual models that match or surpass the capabilities of monolingual models. By leveraging several years of research at Cohere For AI and Cohere, including advancements in data arbitrage, multilingual preference training, and model merging, Aya Expanse sets a new state-of-the-art in multilingual performance. Our evaluations on the Arena-Hard-Auto dataset, translated into 23 languages, demonstrate that Aya Expanse 8B and 32B outperform leading open-weight models in their respective parameter classes, including Gemma 2, Qwen 2.5, and Llama 3.1, achieving up to a 76.6% win-rate. Notably, Aya Expanse 32B outperforms Llama 3.1 70B, a model with twice as many parameters, achieving a 54.0% win-rate. In this short technical report, we present extended evaluation results for the Aya Expanse model family and release their open-weights, together with a new multilingual evaluation dataset m-ArenaHard.

Key facts. Cited by 121 papers in the map (data through 7 Oct 2026). Builds on 34 of its 62 references that are in the map. First posted on arXiv 5 Dec 2024.

Builds on

34 of 62 references.

References · 34

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2022Training language models to follow instructions with human feedbackcited by 15,712
  3. 2024The Llama 3 Herd of Modelscited by 14,192
  4. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  6. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,581
  7. 2024Qwen2 Technical Reportcited by 2,076
  8. 2020GLU Variants Improve Transformercited by 1,724
  9. 2024Gemma: Open Models Based on Gemini Research and Technologycited by 889
  10. 2023Gemini: A Family of Highly Capable Multimodal Modelscited by 810
  11. 2023TIES-Merging: Resolving Interference When Merging Modelscited by 795
  12. 2023AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedbackcited by 738
22 more
  1. 2023Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunchcited by 679
  2. 2023Prometheus: Inducing Fine-grained Evaluation Capability in Language Modelscited by 558
  3. 2021The FLORES-101 Evaluation Benchmark for Low-Resource and Multilingual Machine Translationcited by 556
  4. 2024From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipelinecited by 484
  5. 2023Towards Measuring the Representation of Subjective Global Opinions in Language Modelscited by 424
  6. 2023Scalable Extraction of Training Data from (Production) Language Modelscited by 415
  7. 2023Low-Resource Languages Jailbreak GPT-4cited by 290
  8. 2021Few-shot Learning with Multilingual Language Modelscited by 278
  9. 2023Multilingual Jailbreak Challenges in Large Language Modelscited by 222
  10. 2024Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluationcited by 181
  11. 2024Aya Dataset: An Open-Access Collection for Multilingual Instruction Tuningcited by 174
  12. 2024INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledgecited by 78
  13. 2024The Multilingual Alignment Prism: Aligning Global and Local Preferences to Reduce Harmcited by 67
  14. 2024What Matters for Model Merging at Scale?cited by 52
  15. 2023Privacy in Large Language Models: Attacks, Defenses and Future Directionscited by 48
  16. 2023AfroBench: How Good are Large Language Models on African Languages?cited by 43
  17. 2023GPTAraEval: A Comprehensive Evaluation of ChatGPT on Arabic NLPcited by 37
  18. 2023Lost in Translation: Large Language Models in Non-English Content Analysiscited by 34
  19. 2023On Evaluating and Mitigating Gender Biases in Multilingual Settingscited by 31
  20. 2023A Material Lens on Coloniality in NLPcited by 19
  21. 2023Indian-BhED: A Dataset for Measuring India-Centric Biases in Large Language Modelscited by 17
  22. 2024Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progresscited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 83; 2026: 36.

121 papers.

Sort:

By area:

  1. 2025MultiBLiMP 1.0: A Massively Multilingual Benchmark of Linguistic Minimal Pairscited by 65
  2. 2025Rope to Nope and Back Again: A New Hybrid Attention Strategycited by 46
  3. 2025Multilingual Machine Translation with Open Large Language Models at Practical Scale: An Empirical Studycited by 38
  4. 2025Enhancing LLM Reasoning with Iterative DPO: A Comprehensive Empirical Investigationcited by 36
  5. 2025How Reliable is Multilingual LLM-as-a-Judge?cited by 29
  6. 2025BenchMAX: A Comprehensive Multilingual Evaluation Suite for Large Language Modelscited by 23
  7. 2025Investigating Language Preference of Multilingual RAG Systemscited by 22
  8. 2025BOUQuET: dataset, Benchmark and Open initiative for Universal Quality Evaluation in Translationcited by 21
  9. 2025Localized Cultural Knowledge is Conserved and Controllable in Large Language Modelscited by 19
  10. 2025SEA-HELM: Southeast Asian Holistic Evaluation of Language Modelscited by 17
  11. 2025Commonsense Reasoning in Arab Culturecited by 16
  12. 2025Command R7B Arabic: A Small, Enterprise Focused, Multilingual, and Culturally Aware Arabic LLMcited by 16
38 more
  1. 2025Fann or Flop: A Multigenre, Multiera Benchmark for Arabic Poetry Understanding in LLMscited by 15
  2. 2025CULEMO: Cultural Lenses on Emotion -- Benchmarking LLMs for Cross-Cultural Emotion Understandingcited by 14
  3. 2025Zero-shot Benchmarking: A Framework for Flexible and Scalable Automatic Evaluation of Language Modelscited by 14
  4. 2025Krikri: Advancing Open Large Language Models for Greekcited by 13
  5. 2025NileChat: Towards Linguistically Diverse and Culturally Aware LLMs for Local Communitiescited by 13
  6. 2025Where to show Demos in Your Prompt: A Positional Bias of In-Context Learningcited by 13
  7. 2025DETree: DEtecting Human-AI Collaborative Texts via Tree-Structured Hierarchical Representation Learningcited by 12
  8. 2025SemEval-2025 Task 7: Multilingual and Crosslingual Fact-Checked Claim Retrievalcited by 11
  9. 2025BBQ-V: Benchmarking Visual Stereotype Bias in Large Multimodal Modelscited by 10
  10. 2025The Multilingual Divide and Its Impact on Global AI Safetycited by 10
  11. 2025Sacred or Synthetic? Evaluating LLM Reliability and Abstention for Religious Questionscited by 10
  12. 2025TALES: A Taxonomy and Analysis of Cultural Representations in LLM-generated Storiescited by 10
  13. 2026Back into Plato's Cave: Examining Cross-modal Representational Convergence at Scalecited by 10
  14. 2025When Life Gives You Samples: The Benefits of Scaling up Inference Compute for Multilingual LLMscited by 9
  15. 2025The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLMscited by 8
  16. 2025Global PIQA: Evaluating Commonsense Reasoning Across 100+ Languages and Culturescited by 8
  17. 2025Cetvel: A Unified Benchmark for Evaluating Language Understanding, Generation and Cultural Capacity of LLMs for Turkishcited by 7
  18. 2026EuroLLM-22B: Technical Reportcited by 7
  19. 2025From Guidelines to Practice: A New Paradigm for Arabic Language Model Evaluationcited by 6
  20. 2025From KMMLU-Redux to KMMLU-Pro: A Professional Korean Benchmark Suite for LLM Evaluationcited by 6
  21. 2025DCAD-2000: A Multilingual Dataset across 2000+ Languages with Data Cleaning as Anomaly Detectioncited by 5
  22. 2025Batayan: A Filipino NLP benchmark for evaluating Large Language Modelscited by 5
  23. 2025Cross-Lingual Transfer of Cultural Knowledge: An Asymmetric Phenomenoncited by 5
  24. 2025On Generalization across Measurement Systems: LLMs Entail More Test-Time Compute for Underrepresented Culturescited by 5
  25. 2025BALSAM: A Platform for Benchmarking Arabic Large Language Modelscited by 5
  26. 2025Tailored Emotional LLM-Supporter: Enhancing Cultural Sensitivitycited by 5
  27. 2025Alif: Advancing Urdu Large Language Models via Multilingual Synthetic Data Distillationcited by 5
  28. 2025Model-Aware Tokenizer Transfercited by 5
  29. 2026ReasonXL: Shifting LLM Reasoning Language Without Sacrificing Performancecited by 5
  30. 2025SynthDetoxM: Modern LLMs are Few-Shot Parallel Detoxification Data Annotatorscited by 4
  31. 2025MiLiC-Eval: Benchmarking Multilingual LLMs for China's Minority Languagescited by 4
  32. 2025Language Specific Knowledge: Do Models Know Better in X than in English?cited by 4
  33. 2025EXECUTE: A Multilingual Benchmark for LLM Token Understandingcited by 4
  34. 2025Massively Multilingual Adaptation of Large Language Models Using Bilingual Translation Datacited by 4
  35. 2025Group then Scale: Dynamic Mixture-of-Experts Multilingual Language Modelcited by 4
  36. 2025TokSuite: Measuring the Impact of Tokenizer Choice on Language Model Behaviorcited by 4
  37. 2025Centurio: On Drivers of Multilingual Ability of Large Vision-Language Modelcited by 3
  38. 2025Blessing of Multilinguality: A Systematic Analysis of Multilingual In-Context Learningcited by 3

50 most cited of 121 All 121 (JSON)