Reinforcement Learning Finetunes Small Subnetworks in Large Language Models

Re-assign community