Pinned Loading

1. Beyond-Log-Likelihood Beyond-Log-Likelihood Public

   [ICML'26 Spotlight] What is the right loss function for LLM supervised finetuning?

   Python 65

2. RM-R1-UIUC/RM-R1 RM-R1-UIUC/RM-R1 Public

   [ICLR'26] RM-R1: Unleashing the Reasoning Potential of Reward Models

   Python 165 17

3. JUICE JUICE Public

   [ICML'25 Spotlight] Taming Knowledge Conflict in Language Models

   Jupyter Notebook 6

4. verl-project/verl verl-project/verl Public

   verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework

   Python 22.4k 4.2k