Recipe: Entropy Mechanism
Last updated: 06/27/2025.
The Entropy Mechanism of Reinforcement Learning for Large Language Model Reasoning.
🎉News
[2025/05/29] 🎉 Ranked #1 of the day on Huggingface Daily Papers.
[2025/05/29] Released our Paper on arXiv. See here. We provide insights into the entropy mechanism of RL for LLMs and propose two simple yet effective strategies to alleviate the entropy collapse.
✨Getting started
After preparing the training data, for training Qwen2.5-7B on a single node, taking the KL-Cov approach as an example, you can simply run:
cd verl
conda activate your_env
bash recipe/dapo/7b_kl_cov.sh
While for training Qwen2.5-32B on multi nodes, you can run the following commands:
cd verl
conda activate your_env
bash recipe/dapo/32b_kl_cov.sh
📖Introduction
This paper addresses the entropy collapse issue in scaling reinforcement learning (RL) for large language models (LLMs), where policy entropy drops sharply during training, leading to overconfidence and performance saturation. We empirically establish a relationship between entropy ($H$) and performance ($R$): $R=−aexp(H)+b$, showing performance is bottlenecked by entropy exhaustion.
Theoretically, we find entropy changes are driven by the covariance between action probability and logit updates, which correlates with advantage in Policy Gradient methods. High-probability, high-advantage actions reduce entropy, while rare, high-advantage actions increase it. Empirically, the covariance term remains positive, explaining entropy’s monotonic decline. To mitigate this, we propose Clip-Cov and KL-Cov, which restrict updates for high-covariance tokens. These methods effectively prevent entropy collapse, and improve performance.
📃Evaluation
Our method is able to maintain a considerably higher level of entropy throughout training. For example, when the baseline’s entropy reaches a plateau and can no longer be consumed, the KL-Cov method still sustains an entropy level over 10 times higher. Meanwhile, the response length of the policy model steadily increases, and its performance on the test set consistently surpasses that of the baseline. This indicates that our model is able to explore more freely during training, learning better policy through RL.
Method |
AIME24 |
AIME25 |
AMC |
MATH-500 |
OMNI-MATH |
OlympiadBench |
Minerva |
Avg. |
|---|---|---|---|---|---|---|---|---|
Qwen2.5-7B |
||||||||
GRPO |
21.2 |
9.6 |
58.7 |
78.8 |
27.9 |
40.7 |
36.7 |
38.6 |
w. Clip-higher |
18.1 |
11.5 |
56.6 |
79.2 |
29.8 |
43.3 |
40.4 |
38.8 |
w. |
22.1 |
15.8 |
58.2 |
80.4 |
30.5 |
44.1 |
41.1 |
40.4 |
w. |
22.6 |
12.9 |
61.4 |
80.8 |
29.1 |
42.6 |
38.2 |
40.6 |
Qwen2.5-32B |
||||||||
GRPO |
21.8 |
16.2 |
69.7 |
84.2 |
35.2 |
43.6 |
45.5 |
45.8 |
w. Clip-higher |
35.6 |
22.3 |
69.5 |
77.2 |
35.1 |
42.5 |
43.0 |
47.2 |
w. |
32.3 |
22.7 |
67.2 |
87.0 |
42.0 |
57.2 |
46.0 |
50.3 |
w. |
36.8 |
30.8 |
74.5 |
84.6 |
39.1 |
49.0 |
46.3 |
52.2 |
Our two approaches both achieve non-trivial improvements across all benchmarks. Compared to GRPO, our method outperforms it by 2.0% on average for the 7B model and by 6.4% for the 32B model. Moreover, we observe that our method yields more substantial gains on the larger Qwen2.5-32B. Specifically, our method achieves improvements of 15.0% and 14.6% compared to GRPO on the most challenging benchmarks, AIME24 and AIME25, respectively.
🎈Citation
If you find this paper or repo helpful, please cite us.
@article{cui2025entropy,
title={The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models},
author={Cui, Ganqu and Zhang, Yuchen and Chen, Jiacheng and Yuan, Lifan and Wang, Zhi and Zuo, Yuxin and Li, Haozhan and Fan, Yuchen and Chen, Huayu and Chen, Weize and others},
journal={arXiv preprint arXiv:2505.22617},
year={2025}
}
🌻Acknowledgement
We implement our reinforcement learning algorithm extending from verl. We utilize vLLM for inference. Our models are trained primarily on Qwen2.5 family. Our training data is built from DAPO-MATH. Thanks for their great contributions!
📬 Contact
For questions, discussion, or collaboration opportunities, feel free to contact:
Ganqu Cui: cuiganqu@pjlab.org.cn
Yuchen Zhang: yuchen.zhang2003@gmail.com
Jiacheng Chen: jackchan9345@gmail.com
Ning Ding: ningding.cs@gmail.com