# Language Model Alignment

A decision tree reward model fine-tuned based on Gemma-2-27B, used to evaluate the quality of content generated by language models, with outstanding performance on the RewardBench leaderboard.

URM-LLaMa-3.1-8B is an uncertainty-aware reward model designed to enhance the alignment of large language models.

Gemma 2.9B model fine-tuned on the gemma2-ultrafeedback-armorm dataset using the SimPO objective for preference optimization tasks

SimPO is a preference optimization method that eliminates the need for reference reward models, simplifying the traditional RLHF pipeline by directly optimizing language models with preference data.