GRPO
Group Relative Policy Optimization RL method
Episodes
-
#6288: What Happens Between Model CheckpointsExtending reinforcement learning on Olmo 3 Think for 21 days yielded a 20-point leap on IFBench without starting fresh pretraining. -
#4589: The Answer Reflex: Why AI Models Can't Follow InstructionsDeepSeek passes a simple test that stumps GPT and Claude. Why can't Western models stay in character? -
#2177: Skip Fine-Tuning: Shape LLMs With Alignment AloneCan you build a personalized LLM by skipping traditional fine-tuning and using only post-training alignment methods like DPO and GRPO? We break dow...