Reinforcement Learning with Rubric Anchors

TL;DR

This paper extends reinforcement learning with verifiable rewards to open-ended tasks by using a large, structured rubric-based reward system, enabling improved language model performance and more human-like responses.

Contribution

It introduces a novel rubric-based reward framework for RLVR in open-ended tasks, with the largest rubric system to date and demonstrated improvements on benchmarks.

Findings

Achieved +5.2% improvement on open-ended benchmarks with 5K+ samples

Outperformed a 671B model by +2.4% on certain tasks

Enabled fine-grained stylistic control for more human-like responses

Abstract

Reinforcement Learning from Verifiable Rewards (RLVR) has emerged as a powerful paradigm for enhancing Large Language Models (LLMs), exemplified by the success of OpenAI's o-series. In RLVR, rewards are derived from verifiable signals-such as passing unit tests in code generation or matching correct answers in mathematical reasoning. While effective, this requirement largely confines RLVR to domains with automatically checkable outcomes. To overcome this, we extend the RLVR paradigm to open-ended tasks by integrating rubric-based rewards, where carefully designed rubrics serve as structured, model-interpretable criteria for automatic scoring of subjective outputs. We construct, to our knowledge, the largest rubric reward system to date, with over 10,000 rubrics from humans, LLMs, or a hybrid human-LLM collaboration. Implementing rubric-based RL is challenging; we tackle these issues with a clear framework and present an open-sourced Qwen-30B-A3B model with notable gains: 1) With only 5K+ samples, our system improves by +5.2% on open-ended benchmarks (especially humanities), outperforming a 671B DeepSeek-V3 model by +2.4%, while preserving general and reasoning abilities. 2) Our method provides fine-grained stylistic control, using rubrics as anchors to mitigate the "AI-like" tone and produce more human-like, expressive responses. We share key lessons in rubric construction, data selection, and training, and discuss limitations and future releases.