给神经路由求解器做解释，揭示隐藏的决策逻辑。
Two Black Boxes, One Solver: Encoder Probing and Decoder Attribution for Neural Multi-Attribute VRP under Hard-Mask and Recourse Decoders
- 用探针和归因分析打开编码器与解码器两个黑箱。
- 发现专家混合编码器分布式表示约束，递归训练更擅长暴露可行化路径。
- 适合需要可信决策解释的物流调度系统研发者。
针对多属性车辆路径问题（MAVRP）的神经自回归求解器虽能获得竞争力成本，但缺乏每一步决策的可解释性，不利于调度员验证或比较。本文通过统一协议打开编码器与解码器两个互补的黑箱。编码器侧采用线性探针、自发组织度量、基于秩的丰富度指标及干预验证的发现方向分析，刻画潜在表示在图、节点、边层面如何表征约束族。解码器侧运用梯度、积分梯度、DeepLIFT三种归因方法，从溯因、对比最优可行替代方案、反事实（最小输入改变导致动作切换或恢复可行性）三个角度提供解释。解释效果按保真度、集中度、稳定性、合理性与可操作性评分。在六种组合（三种编码器：Attention baseline、Unimp、UnimpMoe；两种解码器：Hard-Mask、Recourse）中，发现图归纳偏置提升表示可预测性与解码器合理性；专家混合编码器以分布式方式表示约束，而非轴对齐；且递归训练机制不仅带来更软掩码，更能生成有用不可行性表征，揭示使不可行变为可行的反事实路径，而硬掩码策略即使输入不可行方案也无法做到这一点。
原文摘要 · Abstract (English)
Neural autoregressive solvers for the Multi-Attribute Vehicle Routing Problem (MAVRP) reach competitive cost but offer no per-step justification, a problem when dispatchers must validate, accept, or compare them. We open two complementary black boxes in one protocol. On the encoder side, linear probes, spontaneous-organization metrics, rank-based richness measures, and discovered-direction analyses with intervention validation characterize how the latent represents constraint families at the graph, node, and edge level. On the decoder side, three attribution methods (gradient, integrated gradients, DeepLIFT) feed three reading angles: abductive, contrastive against the best feasible alternative, and counterfactual (smallest input change that switches the action or restores feasibility). Explanations are scored on fidelity, concentration, stability, sanity, and actionability. Across six variants combining three encoders (Attention baseline, Unimp, UnimpMoe) with two decoders (Hard-Mask, Recourse), we find that graph inductive bias improves both representational predictability and decoder sanity, that the Mixture-of-Experts encoder represents constraints in a distributed rather than axis-aligned way, and that the Recourse training regime, not merely its softer mask, produces policies that represent infeasibility usefully, exposing make-feasible counterfactuals that Hard-Mask policies fail to produce even when fed infeasible alternatives externally.
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。