Niklas Muennighoff

Niklas Muennighoff on Hugging Face Daily Papers: 19 papers, 14 in the top 3 of their day, 1,179 upvotes.

  1. Crosslingual Reasoning through Test-Time Scaling 8 upvotes, #13 of 2025-05-09
  2. ReasonIR: Training Retrievers for Reasoning Tasks 50 upvotes, #4 of 2025-04-30
  3. s1: Simple test-time scaling 97 upvotes, #1 of 2025-02-03
  4. Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Multimodal Models 89 upvotes, #1 of 2024-09-26
  5. OLMoE: Open Mixture-of-Experts Language Models 67 upvotes, #2 of 2024-09-04
  6. OpenDevin: An Open Platform for AI Software Developers as Generalist Agents 62 upvotes, #1 of 2024-07-25
  7. RegMix: Data Mixture as Regression for Language Model Pre-training 24 upvotes, #7 of 2024-07-02
  8. DataComp-LM: In search of the next generation of training sets for language models 33 upvotes, #2 of 2024-06-18
  9. Language models scale reliably with over-training and on downstream tasks 12 upvotes, #7 of 2024-03-14
  10. StarCoder 2 and The Stack v2: The Next Generation 160 upvotes, #1 of 2024-03-01
  11. Generative Representational Instruction Tuning 54 upvotes, #2 of 2024-02-16
  12. Aya Model: An Instruction Finetuned Open-Access Multilingual Language Model 48 upvotes, #1 of 2024-02-13
  13. Aya Dataset: An Open-Access Collection for Multilingual Instruction Tuning 52 upvotes, #1 of 2024-02-12
  14. Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research 66 upvotes, #2 of 2024-02-02
  15. OLMo: Accelerating the Science of Language Models 86 upvotes, #1 of 2024-02-02
  16. Astraios: Parameter-Efficient Instruction Tuning Code Large Language Models 23 upvotes, #3 of 2024-01-02
  17. FinGPT: Large Generative Models for a Small Language 28 upvotes, #4 of 2023-11-13
  18. OctoPack: Instruction Tuning Code Large Language Models 33 upvotes, #2 of 2023-08-15
  19. StarCoder: may the source be with you! 34 upvotes, #1 of 2023-05-11

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.