👋 Hi, I'm @Rexopia

🚀 Projects:

• 🤗 HawkLM-demo: A 316M parameter LLM pre-trained on Redpajama-1T.
• 🤗 HawkLM-Chat-demo: Instruction-tuned version of HawkLM-demo, fine-tuned on OpenOrca.
• CrewForge: A virtual team in your terminal: multiple AI agents discuss and converge, so you don't have to iterate alone.

🧠 Research Interests:

• Reinforcement learning for sparse reward reasoning: boosting reward signal density via latent self-correction trajectories
• Information geometry of synthetic tabular data: Riemannian manifold representations via Fisher Information pullback

📫 Contact Me:

• 📧 Email: ruiji.zhang@outlook.com