Category: Algorithm Optimization

• Make Thinking More Accurate and Extended! The New Reinforcement Learning Algorithm FIPO Arrives
• Peking University Team Optimizes DeepSeek Attention: 4x Speed Increase Without Accuracy Loss
• Is PPO Dead? The Reinforcement Learning Foundation Used by DeepSeek Has Major Flaws!
• Microsoft Proposes GRPO-RoC: Trajectory Quality Filtering is Key to Agentic RL
• DeepSeek-GRPO Importance Weight Design Flaw? Explaining Qwen3's New Reinforcement Learning Algorithm GSPO