Agent-Conditioned Routing: Diverse Reasoning Paths Within One Frozen Model· 代理条件路由:单一冻结模型中的多路径推理
Generating several solutions can expose alternative methods, but additional trajectories often paraphrase one dominant approach. We investigate whether the conditional capacity of a frozen sparse mixture-of-experts (MoE) model can provide such diversity from one problem prompt, without user-assigned roles, prescribed methods, fine-tuning, or model copies. Agent-Conditioned Routing (ACR) assigns $m$ parallel trajectories persistent, untrained shared/private expert-eligibility masks while preserving the learned, input-dependent token router. On a screened 126-question multi-mode mathematics and physics benchmark, with model judges reading the reported derivations in full, we count the number of problems yielding at least two distinct primary methods among soft-correct solutions in four traje
研究单一冻结模型中多路径推理的多样性。
- 核心方法
- 通过代理条件路由(ACR)方法,在冻结的稀疏专家混合(MoE)模型中为并行轨迹分配持久的共享/私有专家资格掩码,同时保留学习到的输入依赖令牌路由器。
- 适合谁读
- 研究者、工程师
- 要解决的问题
- 如何在不进行额外调整或复制模型的情况下,从单一问题提示中生成多样化的解决方案。
- 关键实验
- 在126个数学和物理问题的基准测试中,通过模型评估员阅读完整的推导过程,统计至少有两种不同主要方法的正确解决方案的数量。
- 主要贡献
- 提出了一种新的路由机制,能够在不改变模型参数的情况下,从单一模型中生成多个不同的解决方案路径。
- 意义与局限
- 该研究为生成多样化解决方案提供了一种新的思路,有望在教育、科研等领域提高问题解决的灵活性和创造性。局限在于目前的实验规模较小,且方法的普适性需要进一步验证。