About Me

Jingwei is now a Principal Researcher at the Technology Innovation Institute (TII), UAE, where he leads the Falcon Foundational Models team. He and his colleagues drive the development of the Falcon family of large language models (LLMs), including Falcon-H1, Falcon-Edge, Falcon 3, and Falcon-Mamba. Their work spans the full LLM lifecycle — from novel architecture design and data curation to large-scale training, deployment, and system-level scaling. The team’s mission is to advance the efficiency, scalability, and usability of Falcon LLMs, bridging fundamental research with real-world applications. Jingwei received his PhD (2022) from University of Paris-Saclay, awarded the Plateau de Saclay Doctoral Prize, an MSc (2018) from University of Paris-Saclay, an Engineer degree (2017) from Sorbonne Université, and a BSc from Huazhong University of Science & Technology.

Current Open Positions (2025)
We’re recruiting LLM research interns (and a small number of full-time roles), If you’re interested, please feel free to reach out.

• Research interns - LLMs
• Researcher - LLMs
• AI Engineer - LLMs
• Senior AI Engineer - LLMs
  

• (2026.1) 🔥 The Falcon-H1-Tiny Series is released! A new family of extremely compact yet remarkably powerful language models: 90M English, 90M Coder, 90M Tool/Function Calling, 100M Multilingual, and 90M / 0.6B Reasoning. More insights can be found in our blogpost
• (2026.1) The Learnable Multiplier paper is out, extending the μ-parameterization used in Falcon-H1 with learnable scaling for improved LLM performance, successfully applied in Falcon-H1-Tiny.
• (2025.7) The technical report of Falcon-H1 is released.
• (2025.6) The E2LM competition got accepted by NeurIPS’25, focused on early-stage training evaluations of LLMs. Registration is now open 🚀 The first prize is 6,000 USD🔥
• (2025.5) 🔥 The Falcon-H1 Series is released! Featuring a novel hybrid Transformer–SSM architecture, Falcon-H1 sets a new state of the art (SoTA) across every scale—0.5B, 1.5B, 3B, 7B, and 34B, and outperforms leading Transformer models double its size. Blogpost 🔍
• (2025.5) 🚀 We just released Falcon Edge! A series of powerful, universal, fine-tunable 1.58bit language models, check the blogpost here.
• (2024.12) Falcon 3 is out! Five base models + their Instruct versions: 1B, 3B, 7B, 10B and a more advanced 7B Mamba model, check the blogpost here.
• (2024.10) The technical report for Falcon Mamba 7B is released - check it here.
• (2024.08) Falcon Mamba 7B is released - the first strong attention-free 7B language model. Try it out in HuggingFace, check the blogpost here.

Preprint:

• M. Velikanov, I. Chahed, J. Zuo, D. Eddine Rhaiem, Y. Belkada, H. Hacid: “Learnable Multipliers: Freeing the Scale of Language Model Matrix Layers”, arXiv’26
• J. Zuo, M. Velikanov, I. Chahed, Y. Belkada, D. Eddine Rhaiem, G. Kunsch, et al.: “Falcon-H1: A Family of Hybrid-Head Language Models Redefining Efficiency and Performance”, arXiv’25
• M. Yagoubi, Y. Dahou, B. Mokeddem, Y. Belkada, P. H. Le-Khac, BB. El Amel Boussaha, R. Alami, Jingwei Zuo, D. Marsili, M. Farooq, M. Lalmas, G. Gkioxari, P. Gallinari, P. Torr, H. Hacid: “NeurIPS 2025 E2LM Competition : Early Training Evaluation of Language Models ”, arXiv’25
• J. Zuo, M. Velikanov, D. Eddine Rhaiem, I. Chahed, Y. Belkada, G. Kunsch and H. Hacid: “Falcon Mamba: The First Competitive Attention-free 7B Language Model”, arXiv’24

• firstname dot lastname at tii dot ae
• PO Box: 9639, Masdar City, Abu Dhabi, UAE