探索用于大语言模型中潜在推理的系统1和系统2通信
Exploring System 1 and 2 communication for latent reasoning in LLMs
AI总结：
本研究探讨了大语言模型中潜在推理的双架构通信方法，发现联合微调比增加通道容量更有效，但需设计明确的目标和训练计划以提升推理能力。
摘要：
大语言模型的推理应该存在于单独的模块中，还是单个模型的前向传递和表示空间中？我们研究了双架构潜在推理，其中一个流畅的基底模型通过潜在消息与协处理器交换信息，并测试了两种假设，旨在改进Liu等人（2024）的潜在通信：（H1）增加通道容量；（H2）通过联合微调学习通信。在GPT-2和Qwen-3上匹配的潜在-令牌预算下，H2表现最强，而H1仅获得有限的提升。一个统一的软嵌入基线，即一个具有相同前向传递和共享表示的单一模型，使用相同的潜在-令牌预算，几乎与H2相当，并超越了H1，表明当前的双架构设计大多只是增加了计算量，而不是在推理上产生质的提升。在GSM8K、ProsQA和一个增加分支因子的Countdown压力测试中，扩大潜在-令牌预算超过小值未能提高鲁棒性。潜在分析显示有重叠的子空间，具有有限的专业化，这与弱推理提升一致。我们得出结论，双模型潜在推理在原则上仍具有潜力，但可能需要明确塑造潜在空间的目标和训练计划，以实现算法规划。
英文摘要：
Should LLM reasoning live in a separate module, or within a single model's forward pass and representational space? We study dual-architecture latent reasoning, where a fluent Base exchanges latent messages with a Coprocessor, and test two hypotheses aimed at improving latent communication over Liu et al. (2024): (H1) increase channel capacity; (H2) learn communication via joint finetuning. Under matched latent-token budgets on GPT-2 and Qwen-3, H2 is consistently strongest while H1 yields modest gains. A unified soft-embedding baseline, a single model with the same forward pass and shared representations, using the same latent-token budget, nearly matches H2 and surpasses H1, suggesting current dual designs mostly add compute rather than qualitatively improving reasoning. Across GSM8K, ProsQA, and a Countdown stress test with increasing branching factor, scaling the latent-token budget beyond small values fails to improve robustness. Latent analyses show overlapping subspaces with limited specialization, consistent with weak reasoning gains. We conclude dual-model latent reasoning remains promising in principle, but likely requires objectives and training schedules that explicitly shape latent spaces for algorithmic planning.
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。
