← 研踪 PaperTrailarXiv:2611.11224 · v1
Do Feature Interactions Break Explainability? Provable Explanations for Any-Order Neural Additive Models
首次提交 2026-11-08 16:14
版本更新 2026-11-08 16:14
雷达观测 2026-11-10 12:16
摘要
Post-hoc explanation methods for neural networks are often heuristic and lack provable guarantees. A common alternative to such methods that was studied in recent years is to compute a cardinally-minimal subset of input features that is provably sufficient to determine the model's prediction. For general neural networks, however, finding such explanations is computationally intractable. In contrast, recent work showed that this problem becomes more tractable for a particular family of neural networks, namely neural additive models (NAMs), though standard NAMs are highly restrictive because they exclude feature interactions altogether. In this work, we study the computation of provably minimal and sufficient explanations for neural networks with restricted feature interactions, bridging the gap between fully general neural networks and purely additive models. We first prove that even with constant-order interactions, the problem is $Σ_2^P$-hard, implying that worst-case exponential complexity is unavoidable. We then identify two provably tractable settings: one based on sparsity in the interaction structure and another based on a stronger interaction-aware notion of sufficiency. Under either of these settings, we develop algorithms that compute provable explanations for neural networks with any-order interactions while retaining much of the computational efficiency of additive models. Empirically, we show that these explanations are significantly smaller and faster to obtain than those produced by standard algorithms, while being derived from models that achieve much higher accuracy than standard NAMs. Overall, our results help expand the class of neural networks that admit efficient provable explanations, while clarifying the fundamental role of feature interactions in shaping their complexity.
作者备注：Accepted at NeurIPS 2026. 36 pages, 2 figures
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-08 16:14 · 最新观测版本
Do Feature Interactions Break Explainability? Provable Explanations for Any-Order Neural Additive Models
Post-hoc explanation methods for neural networks are often heuristic and lack provable guarantees. A common alternative to such methods that was studied in recent years is to compute a cardinally-minimal subset of input features that is provably sufficient to determine the model's prediction. For general neural networks, however, finding such explanations is computationally intractable. In contrast, recent work showed that this problem becomes more tractable for a particular family of neural networks, namely neural additive models (NAMs), though standard NAMs are highly restrictive because they exclude feature interactions altogether. In this work, we study the computation of provably minimal and sufficient explanations for neural networks with restricted feature interactions, bridging the gap between fully general neural networks and purely additive models. We first prove that even with constant-order interactions, the problem is $Σ_2^P$-hard, implying that worst-case exponential complexity is unavoidable. We then identify two provably tractable settings: one based on sparsity in the interaction structure and another based on a stronger interaction-aware notion of sufficiency. Under either of these settings, we develop algorithms that compute provable explanations for neural networks with any-order interactions while retaining much of the computational efficiency of additive models. Empirically, we show that these explanations are significantly smaller and faster to obtain than those produced by standard algorithms, while being derived from models that achieve much higher accuracy than standard NAMs. Overall, our results help expand the class of neural networks that admit efficient provable explanations, while clarifying the fundamental role of feature interactions in shaping their complexity.
查看此版本解读 →查看此版本原文 ↗