Nathan Lambert Tracks Open Model Mentions in Research Papers

Nathan Lambert posted that he parsed 500K arXiv AI/ML papers to track open model usage in research. He stated that in 2024 roughly 30 percent of papers mentioned an American open model and 10 percent a Chinese one. Current shares show about 40 percent mentioning a Chinese open LLM and 25-30 percent an American one. The official Qwen account replied, noting steady growth and thanking him for the analysis.