papers/hllm-training-free-grpo/sections/99-acknowledgments.tex

\section*{Acknowledgments}


This research was conducted by Zoo Labs Foundation, a 501(c)(3) non-profit organization dedicated to democratizing AI. We thank the Hanzo Network team for compute infrastructure, the Lux blockchain team for consensus layer integration, and the global community of contributors to the Gym platform. Special thanks to Tencent's youtu-agent team for pioneering the training-free GRPO algorithm, upon which this work builds.

Funding for this research was provided by community donations to zoo.ngo and KEEPER token holders of the Zoo Network DAO.

\bibliographystyle{plain}
\begin{thebibliography}{99}

\bibitem{kaplan2020scaling}
Kaplan, J., McCandlish, S., Henighan, T., Brown, T.B., Chess, B., Child, R., Gray, S., Radford, A., Wu, J. and Amodei, D., 2020. Scaling laws for neural language models. \textit{arXiv preprint arXiv:2001.08361}.

\bibitem{hoffmann2022training}
Hoffmann, J., Borgeaud, S., Mensch, A., Buchatskaya, E., Cai, T., Rutherford, E., Casas, D.D.L., Hendricks, L.A., Welbl, J., Clark, A. and Hennigan, T., 2022. Training compute-optimal large language models. \textit{arXiv preprint arXiv:2203.15556}.

\bibitem{ouyang2022training}
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A. and Schulman, J., 2022. Training language models to follow instructions with human feedback. \textit{Advances in Neural Information Processing Systems}, 35, pp.27730-27744.

\bibitem{kirkpatrick2017overcoming}
Kirkpatrick, J., Pascanu, R., Rabinowitz, N., Veness, J., Desjardins, G., Rusu, A.A., Milan, K., Quan, J., Ramalho, T., Grabska-Barwinska, A. and Hassabis, D., 2017. Overcoming catastrophic forgetting in neural networks. \textit{Proceedings of the National Academy of Sciences}, 114(13), pp.3521-3526.

\bibitem{wei2022chain}
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Xia, F., Chi, E., Le, Q.V. and Zhou, D., 2022. Chain-of-thought prompting elicits reasoning in large language models. \textit{Advances in Neural Information Processing Systems}, 35, pp.24824-24837.

\bibitem{yao2023tree}
Yao, S., Yu, D., Zhao, J., Shafran, I., Griffiths, T.L., Cao, Y. and Narasimhan, K., 2023. Tree of thoughts: Deliberate problem solving with large language models. \textit{arXiv preprint arXiv:2305.10601}.

\bibitem{brown2020language}
Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J.D., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A. and Agarwal, S., 2020. Language models are few-shot learners. \textit{Advances in Neural Information Processing Systems}, 33, pp.1877-1901.

\bibitem{shao2024deepseekmath}
Shao, Z., Wang, P., Zhu, Q., Xu, R., Song, J., Zhang, M., Li, Y., Wu, Y. and Guo, D., 2024. DeepSeekMath: Pushing the limits of mathematical reasoning in open language models. \textit{arXiv preprint arXiv:2402.03300}.

\bibitem{bai2022constitutional}
Bai, Y., Kadavath, S., Kundu, S., Askell, A., Kernion, J., Jones, A., Chen, A., Goldie, A., Mirhoseini, A., McKinnon, C. and Chen, C., 2022. Constitutional AI: Harmlessness from AI feedback. \textit{arXiv preprint arXiv:2212.08073}.

\bibitem{schulman2017proximal}
Schulman, J., Wolski, F., Dhariwal, P., Radford, A. and Klimov, O., 2017. Proximal policy optimization algorithms. \textit{arXiv preprint arXiv:1707.06347}.

\bibitem{rafailov2023direct}
Rafailov, R., Sharma, A., Mitchell, E., Manning, C.D., Ermon, S. and Finn, C., 2023. Direct preference optimization: Your language model is secretly a reward model. \textit{arXiv preprint arXiv:2305.18290}.

\bibitem{meng2024simpo}
Meng, Y., Xie, S.M., Agarwal, A., Zuo, S., Yin, W. and Liu, T., 2024. SimPO: Simple preference optimization with a reference-free reward. \textit{arXiv preprint arXiv:2405.14734}.

\bibitem{lewis2020retrieval}
Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.T., Rocktäschel, T. and Riedel, S., 2020. Retrieval-augmented generation for knowledge-intensive nlp tasks. \textit{Advances in Neural Information Processing Systems}, 33, pp.9459-9474.

\bibitem{li2021prefix}
Li, X.L. and Liang, P., 2021. Prefix-tuning: Optimizing continuous prompts for generation. \textit{arXiv preprint arXiv:2101.00190}.

\bibitem{lester2021power}
Lester, B., Al-Rfou, R. and Constant, N., 2021. The power of scale for parameter-efficient prompt tuning. \textit{arXiv preprint arXiv:2104.08691}.

\bibitem{laskin2022context}
Laskin, M., Wang, L., Oh, J., Parisotto, E., Spencer, S., Steigerwald, R., Strouse, D.J., Hansen, S., Filos, A., Brooks, E. and Gazeau, M., 2022. In-context reinforcement learning with algorithm distillation. \textit{arXiv preprint arXiv:2210.14215}.

\bibitem{greydanus2019hamiltonian}
Greydanus, S., Dzamba, M. and Yosinski, J., 2019. Hamiltonian neural networks. \textit{Advances in Neural Information Processing Systems}, 32.

\bibitem{cranmer2020lagrangian}
Cranmer, M., Greydanus, S., Hoyer, S., Battaglia, P., Spergel, D. and Ho, S., 2020. Lagrangian neural networks. \textit{arXiv preprint arXiv:2003.04630}.

\bibitem{opper2009variational}
Opper, M. and Archambeau, C., 2009. The variational Gaussian approximation revisited. \textit{Neural Computation}, 21(3), pp.786-792.

\bibitem{mcmahan2017communication}
McMahan, B., Moore, E., Ramage, D., Hampson, S. and Arcas, B.A.Y., 2017. Communication-efficient learning of deep networks from decentralized data. \textit{Artificial Intelligence and Statistics}, pp.1273-1282.

\bibitem{salah2019blockchain}
Salah, K., Rehman, M.H.U., Nizamuddin, N. and Al-Fuqaha, A., 2019. Blockchain for AI: Review and open research challenges. \textit{IEEE Access}, 7, pp.10127-10149.

\bibitem{ghodsi2017safetynets}
Ghodsi, Z., Gu, T. and Garg, S., 2017. SafetyNets: Verifiable execution of deep neural networks on an untrusted cloud. \textit{Advances in Neural Information Processing Systems}, 30.

\bibitem{cover1999elements}
Cover, T.M., 1999. Elements of information theory. \textit{John Wiley \& Sons}.

\bibitem{retool}
Liu, Y., Zhang, R., Song, J., Guo, D. and Wang, P., 2024. ReTool: Enhancing Mathematical Reasoning via Reinforcement Learning with Tool Integration. \textit{arXiv preprint arXiv:2409.17145}.

\bibitem{zhou2022large}
Zhou, Y., Muresanu, A.I., Han, Z., Paster, K., Pitis, S., Chan, H. and Ba, J., 2022. Large language models are human-level prompt engineers. \textit{arXiv preprint arXiv:2211.01910}.

\bibitem{tencentyoutu2025}
Tencent Cloud ADP, 2025. Training-Free Group Relative Policy Optimization: A Novel Approach to Efficient LLM Improvement. \textit{arXiv preprint arXiv:2510.08191v1}.

\bibitem{gym2025}
Zoo Labs Foundation, 2025. Gym: Open-Source AI Model Training Platform. \textit{GitHub repository}, \url{https://github.com/zooai/gym}.

\bibitem{hanzo2025}
Hanzo Network, 2025. Decentralized Compute Infrastructure for AI Workloads. \textit{Technical Documentation}, \url{https://hanzo.network}.

\bibitem{lux2025}
Lux Blockchain, 2025. Multi-Consensus Blockchain Architecture with Post-Quantum Cryptography. \textit{Technical Whitepaper}, \url{https://lux.network}.

\bibitem{zoo2025}
Zoo Network, 2025. Decentralized AI/ML Blockchain with Experience-Based Learning. \textit{Technical Documentation}, \url{https://zoo.ngo}.

\bibitem{zoo-per-llm-chains}
Zoo Labs Foundation, 2025. Per-LLM Chains: Unified Training, Research, and Pooled GPU Hosting on Quasar 3.0. \textit{Zoo Whitepaper}, December 2025.

\bibitem{lp-020-quasar-3}
Lux Foundation, 2025. LP-020: Quasar Consensus 3.0. \textit{Lux Improvement Proposal}.

\bibitem{lp-132-quasargpu}
Lux Foundation, 2025. LP-132: QuasarGPU Adapter. \textit{Lux Improvement Proposal}.

\bibitem{lp-133-app-stack}
Lux Foundation, 2025. LP-133: Quasar-Native App Stack. \textit{Lux Improvement Proposal}.

\bibitem{hanzo-ai-chain}
Hanzo AI, 2025. The Hanzo AI Chain: Inference Marketplace and Commerce Layer. \textit{Hanzo Whitepaper}, December 2025.

\end{thebibliography}