arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.01599cs.LGcs.AI
收藏

多重彩票假说:随机稀疏子网络足以满足RLVR

The Multiple Ticket Hypothesis: Random Sparse Subnetworks Suffice for RLVR

Innopolis University, Tartastan, Russia(塔斯塔兰大学)

Israel Adewuyi, Solomon Okibe, Vladmir Ivanov

AI总结:

该研究提出多重彩票假说,表明在RLVR中,通过训练随机稀疏子网络可达到全参数性能,且不同随机掩码的重叠极低,暗示预训练模型包含多个可行的稀疏子网络。

摘要:

彩票彩票假说证明了稀疏子网络可以匹配全模型性能,表明参数冗余。同时,在具有可验证奖励的强化学习(RLVR)中,最近的研究表明更新集中在参数的稀疏子集上,这进一步支持了这种底层冗余。我们研究了利用这种冗余的最简单方法:在极端稀疏性下仅训练随机选择的参数子集。实证上,我们发现仅训练1%的参数在三个模型和两个任务领域中可以匹配或超过全参数RLVR微调性能。此外,不同的随机掩码显示出极小的重叠(≤0.005的Jaccard相似度),但所有都成功,这表明预训练模型包含许多可行的稀疏子网络,而不是一个特权集。我们称之为多重彩票假说。我们通过RLVR中的隐含每一步KL约束来解释这一现象,该约束限制更新到一个低维子空间,使任意稀疏掩码都能成功。

英文摘要:

The Lottery Ticket Hypothesis demonstrated that sparse subnetworks can match full-model performance, suggesting parameter redundancy. Meanwhile, in Reinforcement Learning with Verifiable Rewards (RLVR), recent work has shown that updates concentrate on a sparse subset of parameters, which further lends evidence to this underlying redundancy. We study the simplest possible way to exploit this redundancy: training only a randomly selected subset of parameters at extreme sparsities. Empirically, we find that training just 1\% of parameters matches or exceeds full-parameter RLVR finetuning across 3 models and 2 task domains. Moreover, different random masks show minimal overlap ($\leq 0.005$ Jaccard similarity) and yet all succeed, suggesting pretrained models contain many viable sparse subnetworks rather than one privileged set. We term this the Multiple Ticket Hypothesis. We explain this phenomenon through the implicit per-step KL constraint in RLVR, which restricts updates to a low-dimensional subspace, enabling arbitrary sparse masks to succeed.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑