Code for NeurIPS 2024 paper "Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs"

RewardAnything: Generalizable Principle-Following Reward Models

This project demonstrates the process of fine-tuning the Qwen2.5-3B-Instruct model using GRPO (Generalized Reward Policy Optimization) on the GSM8K dataset.

ICML'25: The Devil Is in the Details: Tackling Unimodal Spurious Correlations for Generalizable Multimodal Reward Models

RewardAnything: Generalizable Principle-Following Reward Models

Code for "Generalizing reward modeling for out-of-distribution preference learning" in ECML PKDD'2024.

This project demonstrates the process of fine-tuning the Qwen2.5-3B-Instruct model using GRPO (Generalized Reward Policy Optimization) on the GSM8K dataset.

[ACL' 25] The official code repository for DG-PRM: Dynamic and Generalizable Process Reward Modeling

Training generalizable reward models / aligned models

An implementation of the Reward predictive modelling approach introduced in "Reward-predictive representations generalize across tasks in reinforcement learning".

The course teaches how to fine-tune Large Language Models using Reinforcement Learning, specifically GRPO (Generalized Reward-Policy Optimization), instead of supervised labels. Core idea Don’t tell the model the “correct answer.”