让大模型在潜空间中并行推理，提升效率与准确性
LLM Latent Reasoning as Chain of Superposition
- 用潜语义词表约束隐藏状态，保持语义连贯
- 通过诱导监督掩码构建紧凑推理链，长度压缩2.7到5.5倍
- 首次实现潜空间多重推理路径叠加，适合数学推理场景
潜空间推理提供了一种比思维链更高效的计算替代方案，但常因分布错位和链定义模糊导致性能下降。理想情况下，潜空间推理应表现为多个推理路径的叠加。为此，我们提出统一框架Latent-SFT，从词元、推理链和学习三个层面解决挑战：首先，定义潜语义词表（Latent-Vocab），将隐藏状态限制在预训练词汇空间内；其次，通过诱导监督掩码（Induction-Supervision Masking）构建潜空间推理链（Latent-Chain），确保语义紧凑且完整；第三，采用带随机Gumbel-Softmax的潜优化（Latent-Optim），引导模型学习泛化解法。实验表明，Latent-SFT在六个数学基准（如GSM8k、AIME24）上持续优于显式微调，推理长度减少2.7至5.5倍。分析证实，该方法有效捕捉了多种推理轨迹的叠加，而非单一路径压缩。
原文摘要 · Abstract (English)
Latent reasoning offers a computation-efficient alternative to Chain-of-Thought but often suffers from performance degradation due to distributional misalignment and ambiguous chain definitions. Ideally, latent reasoning should function as a superposition of multiple reasoning paths. To realize this, we introduce Latent-SFT, a unified framework addressing challenges at three levels: token, chain, and learning. First, we define the Latent-Vocab to constrain hidden states within the pre-trained vocab-space. Second, we construct the Latent-Chain via Induction-Supervision Masking to ensure semantic compactness and sufficiency. Third, we employ Latent-Optim with stochastic Gumbel-Softmax to guide the model toward generalizable solutions. Empirical results demonstrate that Latent-SFT consistently outperforms explicit SFT across six mathematical benchmarks (e.g., GSM8k, AIME24) while achieving a 2.7x to 5.5x reduction in reasoning length. Analysis confirms that our method effectively captures a superposition of diverse reasoning trajectories rather than merely compressing a single path.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。